Very interesting, isn't it? And I am not surprised! TSMC deserves to move up the top 20 semiconductor companies rankings!! It seems that AMD especially needs to really get its act together.
First, to the rankings. Recently, IC Insights released the list of the top 20 semiconductor sales leaders during Q2-09.Source: IC Insights
In this list, there are four fabless semiconductor companies -- Qualcomm, Broadcom, MediaTek and Nvidia in the top 20, and one foundry -- TSMC, perhaps, emphasizing the growing influence of TSMC as well as the fabless semiconductor companies.
AMD slips! Again?
I had written a couple of posts some time back on AMD and Intel, where the former had commented on the EC ruling on Intel, and also how both were at each other's throats, and had asked the question -- how will all of this help the market?
Well, one hopes that AMD will come back very much stronger in the next quarter, despite its uninspiring guidance for 3Q09, saying that it expects its sales to be "up slightly" from 2Q09.
TSMC, Hynix, MediaTek shine
Coming back to the table, the clear movers are TSMC, and no surprises there, as well as Hynix and MediaTek. In fact, with a little better Q3 performance, TSMC could well move up to the third position, overtaking both Texas Instruments and Toshiba.
Look at the last column -- the 2Q09/1Q09 percentage change -- TSMC has grown by a whopping 93 percent! One other thing! TSMC is reportedly eyeing business opportunities in solar photovoltaics and LEDs in a bid to diversify its revenue channels. Should these happen, expect TSMC to move up higher!
The closest to TSMC in terms of growth are Hynix at 40 percent and Qualcomm at 36 percent, respectively. MediaTek, another impressive mover, grew by 20 percent. Of course, there is Samsung as well, with 29 percent growth.
ST, Micron, Nvidia and NXP have done well too! According to IC Insights, Nvidia replaced Fujitsu in the Q2-09 top 20 rankings. And that brings us to the shakers or those who fared poorly.
Fujitsu, AMD, Freescale slide!
I've already touched upon AMD. Fujitsu cited flash memory and automotive device sales to have suffered immensely this quarter. However, it hopes Q3 will be better and said that customer demand was picking up. So, it could well be back in the Top 20 during Q3.
Yet another slip was in store for Freescale. It slipped from 16th position in 2008 to 18th position during Q1-09, and slid further to 20th position in Q2-09. Perhaps, overdependance on automotives has been its undoing.
An interesting statistic from IC Insights -- Fujitsu, with -9 percent and Freescale, with -2 percent growth, were the only two top-20 companies from Q1-09 to register a 2Q09/1Q09 sales decline!
Wonderful industry guidance
It is heartening to see 19 of the 20 companies registering positive growth this quarter. It won't be improper here to commend IC Insights on its wonderful industry guidance!
In an IC Insights study from late December 2008, it was very vocal in advising firms to adopt a quarterly outlook! It also forecast a significant rebound in the IC market beginning in the third quarter of the year!
IC Insights also stood out by pointing out in early July that H2-09 is likely to usher in strong seasonal strength for electronic system sales, a period of IC inventory replenishment, which began in 2Q09, and positive worldwide GDP growth.
IC Insights had marked 4Q08 as the beginning of the downturn/collapse and Q1-09 as the bottom of the cycle. This quarter (Q2) has largely been a replenishment phase for the inventories. Going by that count, Q3 could well see a true seasonal increase in demand. IC Insights also said that during Q4-09, market growth will mirror the health of the worldwide economy and electronic system sales.
There is light, after all, at the end of the tunnel! Wonder why are the industry folks continue to tell each other -- we still aren't having a good time! Maybe, it is time for them to shed their pessimism and from holding back on investments, and move on to show steely optimism, and indulge in really aggressive buying and selling! After all, work and progress will happen ONLY if you work!!
Friday, July 31, 2009
Thursday, July 30, 2009
Display driver depression follow flat panel succession!
Recently, Randy Lawson, Senior Analyst, Digital TV and Display Electronics, iSuppli Corp., discussed the application market for large and small LCD panel display driver semiconductors, including consumer, monitor monitor/notebook PC displays, consumer plasma displays and cell phone and portable displays.
The LCD driver semiconductor market took a disastrous turn in the second half of 2008 as the economic downturn kicked into high gear and the entire electronics supply chain suffered unprecedented declines. Now, as the industry enters H2-09 and forecasts prognosticating better times, vendors of these display driver ICs are looking at when they will see the market recover.
Revenues history of display driver market
Going by the revenues history, the display driver market peaked in 2005 in terms of revenues. The year 2008 saw revenues for display driver ICs dip ~$1 billion from 2007 levels.
The economic crisis resulting in large production cutbacks in all panel types was the main cause. Also, the driver IC unit shipments fell ~30 percent in 2H-08, compared to 2007 levels.
There have been various factors limiting revenues -- ASP pressure due to panel price, competition, technology shift, particularly, advancements in multichannel and gate-in-panel technologies.
In the last half of 2008 panel production went dramatically low. Some Taiwanese panel fabs were at 50 percent capacity or lower, said Lawson. This market, in terms of iSuppli, has peaked in terms of revenue outlook. It is a very large market in terms of units.
Tracking 2009 recovery
iSuppli has been tracking the monthly shipments of large panel driver ICs in 2009, a main area to watch for recovery signs. Q4-08 was devastating with over 30 percent drop in shipments. However, the large panel driver IC shipments improved from January onward. Also, the panel fab utilization rates increased. The low inventories of IC increased the orders.
However, according to iSuppli, the Q3 outlook is likely to be flat to Q2-09 due to higher quarterly baseline.
Dec. 08 vs. Nov. 08 was down 40 percent in terms of unit shipments. From Jan. 09 onward, shipments started going back up. It really went up in February and March as well. Going into April, things are slowing down a little bit, but it is positive for now. Lawson said that Q3 will likely be pretty flat. The industry is still down on a YoY basis, a point to be noted.
Driver IC units forecast
According to iSuppli, the large LCD saw ~13 percent CAGR and small LCD ~2 percent CAGR. The overall driver IC unit growth rate is likely to be ~10 percent CAGR from 2008-12. Growth will be due primarily to the large panel applications as mobile displays unit growth limit potential for small panel driver ICs, advised Lawson.
"We still have a pretty robust outlook for driver ICs from 2008-2012. LCD TV growth is remaining. Monitors and notebook PCs continue to show relatively strong growth in the long term trend," he said.
Large panels are where the driver ICs will find its biggest opportunity. Small panels will be down this year due to much lower unit shipments. This is due to the quite lower volume shipments of mobile handsets, which make up approximately two-thirds of all categories of drivers in the small categories.
Display driver market forecast -- revenue outlook
In this area, the revenues are likely to be more dictated by large panels. The small panel driver revenues are falling due to the ASP erosion exceeding units growth.
As for the large LCD driver IC revenue swings during the forecast period, 2008 and 2009 will contract due to the overall poor economy hurting customer demand. However, 2010 and 2011 should see strong growth return based on very attractive prices for panels and emerging markets taking more share of LCD TV market and growing.
On the whole, the total revenues are likely to contract >13 percent from 2008 to 2012. The year 2009 will be dramatically down by 20 percent over 2008. "Revenue growth is not there for small LCD drivers. The unit growth strong enough in small drivers to counteract the ASP erosion," said Lawson.
Also, some of the market for small panels is LTPS, which typically has a smaller driver IC and cheaper driver IC anyway, as some of the functionality of the LTPS panels can be integrated into the panel, making for a cheaper driver IC.
Revenue rebound likely in 2010
Definitely, turbulent revenues lie ahead! As mentioned, 2009 driver IC revenues will show significant decline in 2009 over 2008. Panel production levels are still below a year ago levels.
A rebound is likely in 2010, but it won't take the industry back to where it was! Keep in mind that the rebound that happens will be due to a rebound in consumer demand as well as the strength of the China market.
Driver IC unit growth has been slowing in the large panel category. This is due to the adoption of multichannel, high-column drivers as well as the gate-in-panel technology effect. Some maturing in LCD monitor and TV applications in Western markets is also causing slower end system unit growth.
As for small panels, the application growth rate is limited. As mentioned, the cell phone unit growth has been declining. Also, the LTPS share has been growing (driver ICs are smaller and less complex).
There have been continual ASP declines. Also, small panels are transitioning from 130nm to 110nm and 90nm, while large panels transitioning from 0.35um to 0.18um/0.16um. Also, there is a transition from 8-inch to 12-inch wafers.
Market share rankings
In this area, there haven't that many changes. Himax has moved up a bit. iSuppli has added several other companies, such as Lusam, Raydium, Sitronix, Orise, etc., into its tracker.
Q1-09 display driver IC market shares
Q1-09 revenue levels dipped well below Q4-08 revenue levels due to production cutbacks and weak demand in large panel category. The revenue levels were down ~50 percent YoY for Q1-09 as the LCD panel market struggled to find stability in the middle of a disastrous Q4-08 and severe cutbacks in panel production, and thus, IC orders.
The Q4-08 revenues were $1,310mn and Q1-09 revenues were $1,017 revenues -- about 30 percent down. Just for the sake of statistics, Q1-08 was $1,936mn.
Large LCD driver market outlook
Here, revenues are likely to grow ~10 percent over the next four years, primarily a rebound from the dismal H2-08 and 2009 levels. LCD TVs will remain a growth engine, overtaking monitor driver IC volumes from Q3-09 forward.
Lawson said: "There are still large markets such as China, and regions that are still in transition to flat panels. More consumers are buying more TVs per household, and decreasing the time between buying TVs."
However, the monitor driver market has stayed mainly flat, and multichannel use and gate-in-panel are causing diminished unit growth as well.
LCD TV driver type forecast
There is clearly a trend toward multichannel, which is likely to grow for cost, space and reliability savings. It lowers the ICs per panel ratio, and lowers IC unit growth rate as well.
Growth is also expected in 10-bit as well as in TV space for improved image quality. The 10-bit growth may be trimmed due to short term focus on 120/240Hz performance, advised Lawson.
Most of this will be driven by the larger panel category. One trend is toward LED backlighting, which is increasing the contrast ratio.
Small/medium markets scenario
A majority of this market is driven by mobile handsets, which account for two thirds of the market. Almost all of the small display drivers are single chip except for clamshell phones.
Some other major applications of small display drivers include digital cameras/camcorders, automotive, digital photo frames, handheld games/PDAs, PNDs, PMPs, as well as some other applications.
Small display driver forecast
Here, active matrix remains the only growth area left. "Gains will be offset by steep CSTN/MSTN declines," said Lawson. "The ASP erosion is also an issue." There is a transition from 130nm/110nm to 90nm, as well as a move to 300mm and LTPS growth.
All of these present further revenue downward pressure as ASP decline exceed the end market growth. Also, the growth in active matrix is not yet enough to offset the revenue decline in cheap drivers used in CSTN/MSTN.
OLED driver ICs
OLED driver ICs are likely to see revenue growth to ~$150mn by 2012. The OLED volumes are still dominated by mobile communication sub-displays, said Lawson.
The AMOLED move to main displays is likely to be next largest market. It is also getting boost from a market shift from pure-music MP3 players to PMPs.
However, competition from TFT-LCD in the near term is likely to slow the OLED panel volume growth. Also, growth in TVs could be a huge market catalyst, but that is not likely to happen until at least beyond 2011.
Lawson added: "The top line is relatively small as OLEDs are still in a mode of competing against amorphous silicon, LTPS, TFT LCDs, etc.. Until OLEDs can resolve some of the manufacturing issues to get to larger sizes, this will remain one of the smaller markets in the whole driver IC category."
Driver IC process node migration -- 2009-10
A majority are currently built on 0.35micron. As mentioned, a process node transition has been occurring in the driver IC market.
Small/portable display driver designs are leading transition to smallest geometry nodes (90nm from 2010 onward).
Technology trends
As for the technology trends for driver ICs, these include TCON (timing controller) functional integration. Frame rate conversion/MEMC, DisplayPort and other new interfaces becoming more prevalent as well.
In packaging, there is a transition to higher pin count/more channels to save cost (large panel). There is an increased use of COG (chip-on-glass) gate-in-panel technology in monitors/notebooks. These are targeted at smaller screens.
There is also a bit-depth move to 10bit. Here, LCD TV performance is spurred by the DeepColor standard and premium TVs. Another interesting feature is the addition of integrated memory for mobile handset displays.
As for interfaces, handset interfaces will see more of serial high speed to reduce cost, EMI, and interconnectors. Here, there will be a role to play for MIPI (GSM market) and MDDI (CDMA market).
MIPI will probably become the predominant industry standard, but it has yet to take off. MDDI, a Qualcomm standard, has already been deployed in millions of displays.
Next, large panel ICs will move beyond RSDS, LVDS, etc., such as Cascade type, Vx1, PPDS, AiPi, and others. Also, DisplayPort is potentially applicable for panels.
Summary
In summary, the panel driver IC market unit growth is solid, but its revenue presents a different story. The unit volumes growth looks healthy due to primarily large panel applications.
The year 2009 will see improvement from the Q1-09 trough, but it will still not show revenue growth from 2008 level. The ASP erosion is likely to lessen in 2009 due to shortages, but the long term trend remains downward in face of process migration and panel price declines.
Expect more diversification/collaboration from DDI firms, noted Lawson. As the DDI market matures, large companies will be seeking new growth markets, synergy to cut costs and improve efficiencies.
Recent examples include Novatek and Cheertek, Himax Media Solutions and Renesas SP Drivers (Renesas, Powerchip, Sharp combined efforts).
Next, the small LCD and OLED driver markets are dominated by mobile phones. Here, MSTN and CSTN volumes will shrink to less than 15 percent by 2012. Also, TFT, LTPS and OLED will remain the growth areas for small/portable displays.
Finally, the large LCD driver market will see growth in gate-in-panel in LCD monitors, and especially in notebook panels. There will be multichannel, fewer driver ICs per panel, higher reliability, and lower component count. Transition to multichannel is key, as TV transition will impact volumes significantly.
The LCD driver semiconductor market took a disastrous turn in the second half of 2008 as the economic downturn kicked into high gear and the entire electronics supply chain suffered unprecedented declines. Now, as the industry enters H2-09 and forecasts prognosticating better times, vendors of these display driver ICs are looking at when they will see the market recover.
Revenues history of display driver market
Going by the revenues history, the display driver market peaked in 2005 in terms of revenues. The year 2008 saw revenues for display driver ICs dip ~$1 billion from 2007 levels.
The economic crisis resulting in large production cutbacks in all panel types was the main cause. Also, the driver IC unit shipments fell ~30 percent in 2H-08, compared to 2007 levels.
There have been various factors limiting revenues -- ASP pressure due to panel price, competition, technology shift, particularly, advancements in multichannel and gate-in-panel technologies.
In the last half of 2008 panel production went dramatically low. Some Taiwanese panel fabs were at 50 percent capacity or lower, said Lawson. This market, in terms of iSuppli, has peaked in terms of revenue outlook. It is a very large market in terms of units.
Tracking 2009 recovery
iSuppli has been tracking the monthly shipments of large panel driver ICs in 2009, a main area to watch for recovery signs. Q4-08 was devastating with over 30 percent drop in shipments. However, the large panel driver IC shipments improved from January onward. Also, the panel fab utilization rates increased. The low inventories of IC increased the orders.
However, according to iSuppli, the Q3 outlook is likely to be flat to Q2-09 due to higher quarterly baseline.
Dec. 08 vs. Nov. 08 was down 40 percent in terms of unit shipments. From Jan. 09 onward, shipments started going back up. It really went up in February and March as well. Going into April, things are slowing down a little bit, but it is positive for now. Lawson said that Q3 will likely be pretty flat. The industry is still down on a YoY basis, a point to be noted.
Driver IC units forecast
According to iSuppli, the large LCD saw ~13 percent CAGR and small LCD ~2 percent CAGR. The overall driver IC unit growth rate is likely to be ~10 percent CAGR from 2008-12. Growth will be due primarily to the large panel applications as mobile displays unit growth limit potential for small panel driver ICs, advised Lawson.
"We still have a pretty robust outlook for driver ICs from 2008-2012. LCD TV growth is remaining. Monitors and notebook PCs continue to show relatively strong growth in the long term trend," he said.
Large panels are where the driver ICs will find its biggest opportunity. Small panels will be down this year due to much lower unit shipments. This is due to the quite lower volume shipments of mobile handsets, which make up approximately two-thirds of all categories of drivers in the small categories.
Display driver market forecast -- revenue outlook
In this area, the revenues are likely to be more dictated by large panels. The small panel driver revenues are falling due to the ASP erosion exceeding units growth.
As for the large LCD driver IC revenue swings during the forecast period, 2008 and 2009 will contract due to the overall poor economy hurting customer demand. However, 2010 and 2011 should see strong growth return based on very attractive prices for panels and emerging markets taking more share of LCD TV market and growing.
On the whole, the total revenues are likely to contract >13 percent from 2008 to 2012. The year 2009 will be dramatically down by 20 percent over 2008. "Revenue growth is not there for small LCD drivers. The unit growth strong enough in small drivers to counteract the ASP erosion," said Lawson.
Also, some of the market for small panels is LTPS, which typically has a smaller driver IC and cheaper driver IC anyway, as some of the functionality of the LTPS panels can be integrated into the panel, making for a cheaper driver IC.
Revenue rebound likely in 2010
Definitely, turbulent revenues lie ahead! As mentioned, 2009 driver IC revenues will show significant decline in 2009 over 2008. Panel production levels are still below a year ago levels.
A rebound is likely in 2010, but it won't take the industry back to where it was! Keep in mind that the rebound that happens will be due to a rebound in consumer demand as well as the strength of the China market.
Driver IC unit growth has been slowing in the large panel category. This is due to the adoption of multichannel, high-column drivers as well as the gate-in-panel technology effect. Some maturing in LCD monitor and TV applications in Western markets is also causing slower end system unit growth.
As for small panels, the application growth rate is limited. As mentioned, the cell phone unit growth has been declining. Also, the LTPS share has been growing (driver ICs are smaller and less complex).
There have been continual ASP declines. Also, small panels are transitioning from 130nm to 110nm and 90nm, while large panels transitioning from 0.35um to 0.18um/0.16um. Also, there is a transition from 8-inch to 12-inch wafers.
Market share rankings
In this area, there haven't that many changes. Himax has moved up a bit. iSuppli has added several other companies, such as Lusam, Raydium, Sitronix, Orise, etc., into its tracker.
Q1-09 display driver IC market shares
Q1-09 revenue levels dipped well below Q4-08 revenue levels due to production cutbacks and weak demand in large panel category. The revenue levels were down ~50 percent YoY for Q1-09 as the LCD panel market struggled to find stability in the middle of a disastrous Q4-08 and severe cutbacks in panel production, and thus, IC orders.
The Q4-08 revenues were $1,310mn and Q1-09 revenues were $1,017 revenues -- about 30 percent down. Just for the sake of statistics, Q1-08 was $1,936mn.
Large LCD driver market outlook
Here, revenues are likely to grow ~10 percent over the next four years, primarily a rebound from the dismal H2-08 and 2009 levels. LCD TVs will remain a growth engine, overtaking monitor driver IC volumes from Q3-09 forward.
Lawson said: "There are still large markets such as China, and regions that are still in transition to flat panels. More consumers are buying more TVs per household, and decreasing the time between buying TVs."
However, the monitor driver market has stayed mainly flat, and multichannel use and gate-in-panel are causing diminished unit growth as well.
LCD TV driver type forecast
There is clearly a trend toward multichannel, which is likely to grow for cost, space and reliability savings. It lowers the ICs per panel ratio, and lowers IC unit growth rate as well.
Growth is also expected in 10-bit as well as in TV space for improved image quality. The 10-bit growth may be trimmed due to short term focus on 120/240Hz performance, advised Lawson.
Most of this will be driven by the larger panel category. One trend is toward LED backlighting, which is increasing the contrast ratio.
Small/medium markets scenario
A majority of this market is driven by mobile handsets, which account for two thirds of the market. Almost all of the small display drivers are single chip except for clamshell phones.
Some other major applications of small display drivers include digital cameras/camcorders, automotive, digital photo frames, handheld games/PDAs, PNDs, PMPs, as well as some other applications.
Small display driver forecast
Here, active matrix remains the only growth area left. "Gains will be offset by steep CSTN/MSTN declines," said Lawson. "The ASP erosion is also an issue." There is a transition from 130nm/110nm to 90nm, as well as a move to 300mm and LTPS growth.
All of these present further revenue downward pressure as ASP decline exceed the end market growth. Also, the growth in active matrix is not yet enough to offset the revenue decline in cheap drivers used in CSTN/MSTN.
OLED driver ICs
OLED driver ICs are likely to see revenue growth to ~$150mn by 2012. The OLED volumes are still dominated by mobile communication sub-displays, said Lawson.
The AMOLED move to main displays is likely to be next largest market. It is also getting boost from a market shift from pure-music MP3 players to PMPs.
However, competition from TFT-LCD in the near term is likely to slow the OLED panel volume growth. Also, growth in TVs could be a huge market catalyst, but that is not likely to happen until at least beyond 2011.
Lawson added: "The top line is relatively small as OLEDs are still in a mode of competing against amorphous silicon, LTPS, TFT LCDs, etc.. Until OLEDs can resolve some of the manufacturing issues to get to larger sizes, this will remain one of the smaller markets in the whole driver IC category."
Driver IC process node migration -- 2009-10
A majority are currently built on 0.35micron. As mentioned, a process node transition has been occurring in the driver IC market.
Small/portable display driver designs are leading transition to smallest geometry nodes (90nm from 2010 onward).
Technology trends
As for the technology trends for driver ICs, these include TCON (timing controller) functional integration. Frame rate conversion/MEMC, DisplayPort and other new interfaces becoming more prevalent as well.
In packaging, there is a transition to higher pin count/more channels to save cost (large panel). There is an increased use of COG (chip-on-glass) gate-in-panel technology in monitors/notebooks. These are targeted at smaller screens.
There is also a bit-depth move to 10bit. Here, LCD TV performance is spurred by the DeepColor standard and premium TVs. Another interesting feature is the addition of integrated memory for mobile handset displays.
As for interfaces, handset interfaces will see more of serial high speed to reduce cost, EMI, and interconnectors. Here, there will be a role to play for MIPI (GSM market) and MDDI (CDMA market).
MIPI will probably become the predominant industry standard, but it has yet to take off. MDDI, a Qualcomm standard, has already been deployed in millions of displays.
Next, large panel ICs will move beyond RSDS, LVDS, etc., such as Cascade type, Vx1, PPDS, AiPi, and others. Also, DisplayPort is potentially applicable for panels.
Summary
In summary, the panel driver IC market unit growth is solid, but its revenue presents a different story. The unit volumes growth looks healthy due to primarily large panel applications.
The year 2009 will see improvement from the Q1-09 trough, but it will still not show revenue growth from 2008 level. The ASP erosion is likely to lessen in 2009 due to shortages, but the long term trend remains downward in face of process migration and panel price declines.
Expect more diversification/collaboration from DDI firms, noted Lawson. As the DDI market matures, large companies will be seeking new growth markets, synergy to cut costs and improve efficiencies.
Recent examples include Novatek and Cheertek, Himax Media Solutions and Renesas SP Drivers (Renesas, Powerchip, Sharp combined efforts).
Next, the small LCD and OLED driver markets are dominated by mobile phones. Here, MSTN and CSTN volumes will shrink to less than 15 percent by 2012. Also, TFT, LTPS and OLED will remain the growth areas for small/portable displays.
Finally, the large LCD driver market will see growth in gate-in-panel in LCD monitors, and especially in notebook panels. There will be multichannel, fewer driver ICs per panel, higher reliability, and lower component count. Transition to multichannel is key, as TV transition will impact volumes significantly.
Monday, July 27, 2009
How is PV industry reacting to oversupply conditions?
How is the global PV industry reacting to oversupply conditions? Dr. Henning Wicht -- Senior Director and Principal Analyst, iSuppli Corp., and Stefan de Haan -- Senior Analyst, Photovoltaics, iSuppli, have attempted to answer a whole lot of questions and provided an outlook on the global solar photovoltaics industry.
Propelled by exaggerated growth expectations in the past and further boosted by weakening demand due to the global economic downturn, the PV industry has blundered into a situation of critical oversupply becoming evident on all nodes of the solar value chain.
According to Dr. Wicht, last October, all prices of wafers and modules were up, and nobody had foreseen a dramatic drop. It happened indeed, because many people believed in the industry consensus.
Recent findings -- June 09
1. Polysilicon supply will enforce further price drops in 2010
2. You will see mounting inventories - due to mounting inventories, module production in 2009 will be ~25 percent smaller than expected.
3, Module price decline will continue, but 2009 is slower already
4. Global PV installations, though in good shape, wont exceed 4GW in 2009.
PV market update -- key findings
Polysilicon
- Oversupply situation is getting worse
- Significant price drops expected through 2010: $40/kg or less
- Winners will be determined by cost structures
Modules
- Inventory levels climb to three times that of Q2-2008
- Production cuts and review of business model in 2009
- Prices continue to drop, but decline will slow down at the end of 2009: $2.2-2.3/W end 2009
PV system demand
- Global installations 2009: 4GW to be installed this year
- Revenue decline of 35 percent compared to 2008 -- since we have a price decline and volume decline
The solar industry will continue to change, said Dr. Wicht. "Oversupply situation is not the end of the industry. It will pass relatively soon. Suppliers and supplier structures will change. Once dynamics change, the prices will change as well."
Polysilicon prices
According to Stefan de Haan, polysilicon price peaked in 2008. Due to the changing dynamics of supply and demand, we forecast prices will drop significantly soon after 2008. What we see today is indeed this picture. Prices have dropped even further down. Spot prices are in the range of $70-90 per kilogram. It is likely to continue.
"We also forecasted on the price. Demand will settle in somewhere at the 3GW range for crystalline modules. At that level, you'll have suppliers at $2.5, $2.8 per W."
So, the scenario is: price is likely to be in the $2.5 per W range for crystalline modules in 2009. What we see now in the market is that prices are already at $2.7 per W. Q-Cells, for instance, had warned on July 14 that Q2 was not as good as they had expected in Q1.
Market in next 12-14 months
Toucing on the PV market update, Stefan de Haan said: "What iSuppli's model had predicted last year has now become reality. The PV markets shifted from supply-constrained to demand-constrained -- and that's a fact -- in 2009. Companies were forced to react to oversupply situation. There were severe production cuts and unutilized capacity."
These effects are reflected in iSuppli's models.
* Rreal production' vs. 'production roadmap'
* Key element to anticipate short-term strategy changes
* Inventory level analysis
As the mismatch of demand and supply is worsening, so companies are forced to react to this. They not only delayed or cancelled their long term expansion plans, but also significantly reduced their current capacity utilization and really cut back on actual production.
To reflect these strategy shifts in our models, we distinguish between the 'real production' vs. 'production roadmap'.
"The fundamental question we ask is: how likely is that companies will achieve what they plan. If supply and demand become severely disconnected, we also have to consider short-term strategy changes to forecast real production," he added.
Current production indicates a PV modules oversupply of over 6GW -- which will result in $12 billion spent only in building actual module inventory. This number is not possible as the industry will literally go bankrupt with that level of investment.
How much inventory can the industry afford?
The level of 30-50 days of inventories has been standard during Q1-08. Now, in Q4-08, inventories are growing enormously. Only First Solar has managed to maintain a level of less than 30 days inventory (28 days) during Q1-09.
"When you compare First Solar and Solon, the picture becomes really dramatic. When you look at Solon's inventory amounted to $237mn compared to revenues of $50 million -- or 427 days of inventory. Solon was first module producer who dramatically cut back its production. It now disposed of annual capacity of 500MW, but only produced 18MW of modules in Q109," he said.
Many other module suppliers, very similar to Solon, do not have sufficient resources to invest in further inventory. Their cash flow in Q1 is already negative.
In comparable industries it is a rule of thumb that companies cannot afford to exceed inventory levels by three to four times than what the market needs. If you take 28 days inventory of First Solar, many companies, are approaching 120 days or even exceeding that limit.
For the short term forecast, iSuppli tried to determine how strongly the industry should react to current environment. In the module level, we will still see many decisions to scale back output in next quarter.
Polysilicon demand-supply balance
In 2009, solar polysilicon production will grow to 14GW, up from 6.5GW in 2008. The supply will be more than doubling this year. Crystalline cell production will be growing only moderately, from 5.8GW in 2008 to 7GW in 2009.
The polysilicon shortage has already ended last year. At that time, supply and demand were really balanced. We wil see a polysilicon oversupply of almost 100 percent in 2009.
The polysilicon industry reacts to market changes with less flexibility than cell and module producers. The investments are very high, ramp rates are long, and so, it is not so easy to build up polysilicon production facility. The whole industry is just slow in its movements than other parts of the PV supply chain. Another reason for worsening mismatch is that the silicon consumption per W has also decreased faster than previously expected.
July 2009 -- polysilicon price projection
Prices peaked in 2008 -- $350-500 per kg. However, it is clear that since Oct. 2008, prices dropped dramatically. In May 2009, spot average was around $100 per kg. Right now, we are already below $80 and the decline will continue.
Regarding the building oversupply, prices will soon drop to production cost levels. That's how price is built in an oversupply scenario. Companies will do everything to clear their excess production, There will be only room for small margin for best in class producers. In 2010, perhaps, earlier, we may see spot price of $40 per kg.
Module demand-supply balance
Oversupply had already started to build up in 2008. In 2009, 5.8GW of crystalline modules will be produced and 1.7GW of thin film modules will be produced. This gives 7.5GW in total. However, the total installations are 3.9GW, and not more than 4GW. This translates into a module oversupply of about 90 percent this year.
These numbers already include short-term production cuts. They are already heavily discounted. For every module that will be produced in 2009, almost 1 additional module will go into inventory. Companies react to this and adjust their current outputs and long term plans.
Q-Cells slowed down expansion in Malaysia. Sun Power delayed rampup in the Philippines and Malaysia. Moser Baer also delayed its rampup of the new thin film factory.
Stefan de Haan said: "There are many cancellations that are happening now. We will see a decrease in oversupply in the following years. The next three quarters will still be tough for the module suppliers."
Module price projection
* Total c-Si module production in 2009: 6.79GW
* Total c-Si module demand in 2009: 2.50GW
* Oversupply: 132 percent
* Pure play (module manufacturers) -- production 1.37GW, cost $2.41 per W.
* Integrated cell and module manufacturers -- production 2.64GW, cost $2.38 per W.
* Integrated wafer, cell and module manufacturers -- production 1.60GW, cost $2.16 per W.
* Integrated p-Si wager cell and module manufacturers -- production 0.19GW, cost $1.88 per W.
In 2009, c-Si price will reach $2.3 per W. Cost will determine the price. Prices will drop further until the end of this year, but not as rapidly as at Q1.
Last November, iSuppli had predicted that prices will drop from $4 per W to $2.5 per W until the end of 2009. It has now corrected this and expects prices to be $2.3 per W. Production costs have dropped, and hence the reason for this adjustment.
Margins get compressed!
In current environments, margins get compressed along the entire supply chain. Significant price drops have completely eaten up margins of the cell and module producers. Module producers are passing their pressure on. Wafer manufacturers can still maintain very small margins.
"Polysilicon production for incumbent producers is still better, but the situation will worsen in the next year. Winners and losers on module level are not so easy to identify any more, Stefan de Haan added.
When will demand pick up again?
According to Dr, Henning Wicht, total installations in 2009 are expected to be 3.9GW, which is less than 2008, which was 5.5GW. "This installation will again grow in the years ahead. The year-to-year growth will be good, but all of this has to be balanced against supply. We analyzed the regions, which make 80-90 percent of the total installations. The solar industry is changing its mood rapidly."
Installations in major regions -- well, there are no big surprises! The leading regions are: Germany, Italy, Spain, California (USA), France, Greece, Japan, Bulgaria, Czech Republic, and the RoW.
The total capacity in March 2009 was 3,546 MW; which is estimated for June 2009 at 3,922 MW. By 2013, it should be 25,900MW. Also, Italy is probably being stronger than we expected, and Spain is lesser. California is growing faster than expected earlier.
PV installations by market segment
Dr. Wicht said: "Looking at the global annual PV installations by market segment. The market today is dominated by the European countries, mainly Germany. The German market is mainly a rooftop market. We have residential and commercial roofs, which provide 90 percent of the installations in Germany."
In future, ground installations will pick up. This is mainly due to the US and the potential of China, and the new entrants, who will definitely address this segment, more than the others.
Outlook -- short term
In the short term, credits for solar installations remain tight.
* Bankers do not see light at the end of the tunnel.
- Interbank lending is nearly zero.
- Solar installations compete against numerous other investments.
* Despite excellent investment conditions, no installation boom to come.
- Limited cash and credit available.
* Economic downturn has not yet impacted solar investments in Germany.
- Increasing RoI compensates for economic uncertainty
- Private investors can access loan programs
Dr. Wicht said: "The short-term picture is dominated by tight credits! A panel of bankers clearly indicated that they love solar, but don't have the money right now. They have clear expectations on their solar RoI. Now, the portfolio for solar is already booked. So, there's no need for bankers to invest much more than what we have seen."
For instance, in Italy, conditions are very good, but lenders do not accept thin film. Thin film does not have long term experience like crystalline. With crystalline, they already meet their RoI expectations due to lower module prices.
Solar installations are still in good shape. The economic downturn has not yet impacted the solar investments in Germany. Rooftops -- commercial and residential are in good shape. They are likely to achieve the same levels as last year.
Outlook -- mid-term
In the mid term, there are exciting times ahead as the big wave of solar is rising.
* RoW is preparing large solar programs.
- Australia will invest in 1GW; China and India. India is also likely to concretize its plans.
* Every month, new several 100MW announcements are being made in US, especially, California, Nevada. The US is picking up in solar.
* In Europe, the Desertec Group -- Euro 400bn to develop solar plants in North Africa
- Lead by Munchner Ruck (Munich Re), Siemens, Deutsche Bank and RWE
- Big credit to solar (all technologies)
- Equal to equal discussions with uranium, fossil energy providers
The priority will be on solar power plants, but we are certain there is space for solar, PV and PV concentrated installations.
* China will not leave REE business to US and Europeans
Desertec project update
On July 13, 15 industrial members joined the Desertec Interest Group. They made the statement that they want to make the solar power plant to supply 20 percent of the European electricity by 2020 from North Africa.
As for the next step, over the next two years, feasibility studies will be conducted on solar power plants as well as grid connection, and transporting of energy through the whole of Europe.
Propelled by exaggerated growth expectations in the past and further boosted by weakening demand due to the global economic downturn, the PV industry has blundered into a situation of critical oversupply becoming evident on all nodes of the solar value chain.
According to Dr. Wicht, last October, all prices of wafers and modules were up, and nobody had foreseen a dramatic drop. It happened indeed, because many people believed in the industry consensus.
Recent findings -- June 09
1. Polysilicon supply will enforce further price drops in 2010
2. You will see mounting inventories - due to mounting inventories, module production in 2009 will be ~25 percent smaller than expected.
3, Module price decline will continue, but 2009 is slower already
4. Global PV installations, though in good shape, wont exceed 4GW in 2009.
PV market update -- key findings
Polysilicon
- Oversupply situation is getting worse
- Significant price drops expected through 2010: $40/kg or less
- Winners will be determined by cost structures
Modules
- Inventory levels climb to three times that of Q2-2008
- Production cuts and review of business model in 2009
- Prices continue to drop, but decline will slow down at the end of 2009: $2.2-2.3/W end 2009
PV system demand
- Global installations 2009: 4GW to be installed this year
- Revenue decline of 35 percent compared to 2008 -- since we have a price decline and volume decline
The solar industry will continue to change, said Dr. Wicht. "Oversupply situation is not the end of the industry. It will pass relatively soon. Suppliers and supplier structures will change. Once dynamics change, the prices will change as well."
Polysilicon prices
According to Stefan de Haan, polysilicon price peaked in 2008. Due to the changing dynamics of supply and demand, we forecast prices will drop significantly soon after 2008. What we see today is indeed this picture. Prices have dropped even further down. Spot prices are in the range of $70-90 per kilogram. It is likely to continue.
"We also forecasted on the price. Demand will settle in somewhere at the 3GW range for crystalline modules. At that level, you'll have suppliers at $2.5, $2.8 per W."
So, the scenario is: price is likely to be in the $2.5 per W range for crystalline modules in 2009. What we see now in the market is that prices are already at $2.7 per W. Q-Cells, for instance, had warned on July 14 that Q2 was not as good as they had expected in Q1.
Market in next 12-14 months
Toucing on the PV market update, Stefan de Haan said: "What iSuppli's model had predicted last year has now become reality. The PV markets shifted from supply-constrained to demand-constrained -- and that's a fact -- in 2009. Companies were forced to react to oversupply situation. There were severe production cuts and unutilized capacity."
These effects are reflected in iSuppli's models.
* Rreal production' vs. 'production roadmap'
* Key element to anticipate short-term strategy changes
* Inventory level analysis
As the mismatch of demand and supply is worsening, so companies are forced to react to this. They not only delayed or cancelled their long term expansion plans, but also significantly reduced their current capacity utilization and really cut back on actual production.
To reflect these strategy shifts in our models, we distinguish between the 'real production' vs. 'production roadmap'.
"The fundamental question we ask is: how likely is that companies will achieve what they plan. If supply and demand become severely disconnected, we also have to consider short-term strategy changes to forecast real production," he added.
Current production indicates a PV modules oversupply of over 6GW -- which will result in $12 billion spent only in building actual module inventory. This number is not possible as the industry will literally go bankrupt with that level of investment.
How much inventory can the industry afford?
The level of 30-50 days of inventories has been standard during Q1-08. Now, in Q4-08, inventories are growing enormously. Only First Solar has managed to maintain a level of less than 30 days inventory (28 days) during Q1-09.
"When you compare First Solar and Solon, the picture becomes really dramatic. When you look at Solon's inventory amounted to $237mn compared to revenues of $50 million -- or 427 days of inventory. Solon was first module producer who dramatically cut back its production. It now disposed of annual capacity of 500MW, but only produced 18MW of modules in Q109," he said.
Many other module suppliers, very similar to Solon, do not have sufficient resources to invest in further inventory. Their cash flow in Q1 is already negative.
In comparable industries it is a rule of thumb that companies cannot afford to exceed inventory levels by three to four times than what the market needs. If you take 28 days inventory of First Solar, many companies, are approaching 120 days or even exceeding that limit.
For the short term forecast, iSuppli tried to determine how strongly the industry should react to current environment. In the module level, we will still see many decisions to scale back output in next quarter.
Polysilicon demand-supply balance
In 2009, solar polysilicon production will grow to 14GW, up from 6.5GW in 2008. The supply will be more than doubling this year. Crystalline cell production will be growing only moderately, from 5.8GW in 2008 to 7GW in 2009.
The polysilicon shortage has already ended last year. At that time, supply and demand were really balanced. We wil see a polysilicon oversupply of almost 100 percent in 2009.
The polysilicon industry reacts to market changes with less flexibility than cell and module producers. The investments are very high, ramp rates are long, and so, it is not so easy to build up polysilicon production facility. The whole industry is just slow in its movements than other parts of the PV supply chain. Another reason for worsening mismatch is that the silicon consumption per W has also decreased faster than previously expected.
July 2009 -- polysilicon price projection
Prices peaked in 2008 -- $350-500 per kg. However, it is clear that since Oct. 2008, prices dropped dramatically. In May 2009, spot average was around $100 per kg. Right now, we are already below $80 and the decline will continue.
Regarding the building oversupply, prices will soon drop to production cost levels. That's how price is built in an oversupply scenario. Companies will do everything to clear their excess production, There will be only room for small margin for best in class producers. In 2010, perhaps, earlier, we may see spot price of $40 per kg.
Module demand-supply balance
Oversupply had already started to build up in 2008. In 2009, 5.8GW of crystalline modules will be produced and 1.7GW of thin film modules will be produced. This gives 7.5GW in total. However, the total installations are 3.9GW, and not more than 4GW. This translates into a module oversupply of about 90 percent this year.
These numbers already include short-term production cuts. They are already heavily discounted. For every module that will be produced in 2009, almost 1 additional module will go into inventory. Companies react to this and adjust their current outputs and long term plans.
Q-Cells slowed down expansion in Malaysia. Sun Power delayed rampup in the Philippines and Malaysia. Moser Baer also delayed its rampup of the new thin film factory.
Stefan de Haan said: "There are many cancellations that are happening now. We will see a decrease in oversupply in the following years. The next three quarters will still be tough for the module suppliers."
Module price projection
* Total c-Si module production in 2009: 6.79GW
* Total c-Si module demand in 2009: 2.50GW
* Oversupply: 132 percent
* Pure play (module manufacturers) -- production 1.37GW, cost $2.41 per W.
* Integrated cell and module manufacturers -- production 2.64GW, cost $2.38 per W.
* Integrated wafer, cell and module manufacturers -- production 1.60GW, cost $2.16 per W.
* Integrated p-Si wager cell and module manufacturers -- production 0.19GW, cost $1.88 per W.
In 2009, c-Si price will reach $2.3 per W. Cost will determine the price. Prices will drop further until the end of this year, but not as rapidly as at Q1.
Last November, iSuppli had predicted that prices will drop from $4 per W to $2.5 per W until the end of 2009. It has now corrected this and expects prices to be $2.3 per W. Production costs have dropped, and hence the reason for this adjustment.
Margins get compressed!
In current environments, margins get compressed along the entire supply chain. Significant price drops have completely eaten up margins of the cell and module producers. Module producers are passing their pressure on. Wafer manufacturers can still maintain very small margins.
"Polysilicon production for incumbent producers is still better, but the situation will worsen in the next year. Winners and losers on module level are not so easy to identify any more, Stefan de Haan added.
When will demand pick up again?
According to Dr, Henning Wicht, total installations in 2009 are expected to be 3.9GW, which is less than 2008, which was 5.5GW. "This installation will again grow in the years ahead. The year-to-year growth will be good, but all of this has to be balanced against supply. We analyzed the regions, which make 80-90 percent of the total installations. The solar industry is changing its mood rapidly."
Installations in major regions -- well, there are no big surprises! The leading regions are: Germany, Italy, Spain, California (USA), France, Greece, Japan, Bulgaria, Czech Republic, and the RoW.
The total capacity in March 2009 was 3,546 MW; which is estimated for June 2009 at 3,922 MW. By 2013, it should be 25,900MW. Also, Italy is probably being stronger than we expected, and Spain is lesser. California is growing faster than expected earlier.
PV installations by market segment
Dr. Wicht said: "Looking at the global annual PV installations by market segment. The market today is dominated by the European countries, mainly Germany. The German market is mainly a rooftop market. We have residential and commercial roofs, which provide 90 percent of the installations in Germany."
In future, ground installations will pick up. This is mainly due to the US and the potential of China, and the new entrants, who will definitely address this segment, more than the others.
Outlook -- short term
In the short term, credits for solar installations remain tight.
* Bankers do not see light at the end of the tunnel.
- Interbank lending is nearly zero.
- Solar installations compete against numerous other investments.
* Despite excellent investment conditions, no installation boom to come.
- Limited cash and credit available.
* Economic downturn has not yet impacted solar investments in Germany.
- Increasing RoI compensates for economic uncertainty
- Private investors can access loan programs
Dr. Wicht said: "The short-term picture is dominated by tight credits! A panel of bankers clearly indicated that they love solar, but don't have the money right now. They have clear expectations on their solar RoI. Now, the portfolio for solar is already booked. So, there's no need for bankers to invest much more than what we have seen."
For instance, in Italy, conditions are very good, but lenders do not accept thin film. Thin film does not have long term experience like crystalline. With crystalline, they already meet their RoI expectations due to lower module prices.
Solar installations are still in good shape. The economic downturn has not yet impacted the solar investments in Germany. Rooftops -- commercial and residential are in good shape. They are likely to achieve the same levels as last year.
Outlook -- mid-term
In the mid term, there are exciting times ahead as the big wave of solar is rising.
* RoW is preparing large solar programs.
- Australia will invest in 1GW; China and India. India is also likely to concretize its plans.
* Every month, new several 100MW announcements are being made in US, especially, California, Nevada. The US is picking up in solar.
* In Europe, the Desertec Group -- Euro 400bn to develop solar plants in North Africa
- Lead by Munchner Ruck (Munich Re), Siemens, Deutsche Bank and RWE
- Big credit to solar (all technologies)
- Equal to equal discussions with uranium, fossil energy providers
The priority will be on solar power plants, but we are certain there is space for solar, PV and PV concentrated installations.
* China will not leave REE business to US and Europeans
Desertec project update
On July 13, 15 industrial members joined the Desertec Interest Group. They made the statement that they want to make the solar power plant to supply 20 percent of the European electricity by 2020 from North Africa.
As for the next step, over the next two years, feasibility studies will be conducted on solar power plants as well as grid connection, and transporting of energy through the whole of Europe.
Thursday, July 23, 2009
SAP World Tour '09: Heralding a "Clear Enterprise"
Challenges that businesses face today are still around! Although India seems to have seen the worst of the economic downturn, contends Ranjan Das, managing director, SAP Indian Subcontinent, SAP India Pvt Ltd, business as usual, as we know, is now dead. The reason being, there's lot more volatility and risk present. There are also a lot of regulations. Besides, there exists a gap between strategy and execution.
According to a survey, over 60 percent of businesses are still thinking about laying off people in 2009. About 40 percent of the businesses are thinking about restructuring. About a third of the businesses are looking at introducing new products, and all of them are looking at cost cutting!
What's a clear enterprise?
It would be interesting to see what the well-run companies are doing!
1. They have an enhanced focus on customers.
2. They are focused on getting more out of employees and also that the employees themselves are growing.
3. Good companies are focusing on cutting costs and enhancing their brands.
SAP, which has 80,000+ customers globally, and 4,000+ customers in India alone, has come up with the idea of a "Clear Enterprise." Enterprises need to think, see and execute clearly! Clear enterprises need to do the following:
a) They are agile!
b) They are collaborative -- within enterprises and across the entire value chain.
c) Focus on transparency and accountability.
d) Focus on sustainability.
Perhaps, I would agree with all of these points, but then, these have been talked about for a long time now! An agile enterprise is bound to have all of these qualities. So, am not quite sure whether 'clear' is any different than 'agile'! Nevertheless!!
Citing an example of BSNL, India's leading telco, Das said that until recently, BSNL did not really have enterprise visibility. "They realized it, and are now trying to get closer to its customers and also have a clear visibility of assets," he added. "It is in the path toward becoming a clear enterprise."
Commenting on SAP's current status, Das noted that the company is in a transformation phase toward becoming a clear enterprise. The company had re-organized itself to achieve this goal. SAP has especially cut down on the carbon footprint. It has also acquired Clear Standards.
A privately held innovator of enterprise carbon management solutions, Clear Standards helps organizations accurately measure, optimize and report greenhouse gas (GHG) emissions and other environmental impacts across internal operations.
With this move, SAP expects to accelerate its ability to meet the carbon management requirements of organizations in this time of increasingly stringent government regulations and expectations for better transparency by the public.
Clear Standards provides SAP a mature sustainability solution and expertise in carbon management delivered through an agile, Web-based, on-demand delivery model. In order to reduce customer effort and cost around carbon management, SAP will leverage its business process expertise to enable Clear Standards to tap into financial and other data stored in enterprise solutions such as SAP Business Suite 7 software and the SAP Environment, Health, and Safety Management application.
Das further added that SAP looked at social accountability and corporate social responsibility (CSR) as well. These include: economic bottomline, environmental bottomline and social bottomline, respectively.
SAP has 75 percent SMEs as customers and the enterprises make up the remaining 25 percent. Reliance, ONGC, etc., are among its major Indian customers. Das said, "We enable customers to do three things -- help automate, and if automated, optimize; provide companies with best practices -- and Indian companies are certainly very interested in best practices; and enable businesses/users at all levels to make decisions."
He added that ERP is a subset of all processes. In this respect, SAP is much more than an ERP company. Commenting on Business ByDesign, he added that it has been introduced in six countries. "When we are ready, India will be one of the key markets. It will help a lot of customers, especially, the SMEs, to grow," he added.
SAP BusinessObjects Explorer
Sanjay Deshmukh, vice president, Business User and Platforms, SAP India Pvt Ltd, took the opportunity to introduce the SAP BusinessObjects Explorer.Sanjay Deshmukh, vice president, Business User and Platforms, SAP India Pvt Ltd, making a point, while Ranjan Das, managing director, SAP Indian Subcontinent, SAP India Pvt Ltd, looks on.
This tool -- the SAP BusinessObjects Explorer software -- enables you to put the power of business intelligence (BI) into the hands of all of your business users, arming any employee with the clarity and insight to act more quickly and make smarter, more effective business decisions.
According to Deshmukh, a majority of the casual users are underserved by traditional BI tools. A majority of users tend to use search engines while looking for information. The SAP BusinessObjects Explorer software makes use of the power of the Web as well!
This solution provides intuitive experience, data and speed -- rather, immediate insight into mountains of data! "Explorer is a self service, search driven and discovery solution for exploring and analyzing your corporate data," added Deshmukh. "It brings BI to all business users and helps IT to become successful." Among SAP APJ's ramp-up customers are names such as Mahindra and Innogence.
He estimated the Indian BI market to be worth $75-100 million, and about 25-30 percent market share -- the highest market share as per analysts.
Deshmukh added that as a standalone entity, BusinessObjects had the largest share in BI. "Since being bought by SAP, we have also provided clear, futuristic view landscape for customers. Second, it helped us understand business processes better," he said. "Third, SAP also brought in a lot of solutions, such as EPMS, GRC, etc. BI is just a stepping stone toward becoming a clear enterprise."
Before I end, may I take this opportunity to thank SAP for inviting me to this wonderful conference!
According to a survey, over 60 percent of businesses are still thinking about laying off people in 2009. About 40 percent of the businesses are thinking about restructuring. About a third of the businesses are looking at introducing new products, and all of them are looking at cost cutting!
What's a clear enterprise?
It would be interesting to see what the well-run companies are doing!
1. They have an enhanced focus on customers.
2. They are focused on getting more out of employees and also that the employees themselves are growing.
3. Good companies are focusing on cutting costs and enhancing their brands.
SAP, which has 80,000+ customers globally, and 4,000+ customers in India alone, has come up with the idea of a "Clear Enterprise." Enterprises need to think, see and execute clearly! Clear enterprises need to do the following:
a) They are agile!
b) They are collaborative -- within enterprises and across the entire value chain.
c) Focus on transparency and accountability.
d) Focus on sustainability.
Perhaps, I would agree with all of these points, but then, these have been talked about for a long time now! An agile enterprise is bound to have all of these qualities. So, am not quite sure whether 'clear' is any different than 'agile'! Nevertheless!!
Citing an example of BSNL, India's leading telco, Das said that until recently, BSNL did not really have enterprise visibility. "They realized it, and are now trying to get closer to its customers and also have a clear visibility of assets," he added. "It is in the path toward becoming a clear enterprise."
Commenting on SAP's current status, Das noted that the company is in a transformation phase toward becoming a clear enterprise. The company had re-organized itself to achieve this goal. SAP has especially cut down on the carbon footprint. It has also acquired Clear Standards.
A privately held innovator of enterprise carbon management solutions, Clear Standards helps organizations accurately measure, optimize and report greenhouse gas (GHG) emissions and other environmental impacts across internal operations.
With this move, SAP expects to accelerate its ability to meet the carbon management requirements of organizations in this time of increasingly stringent government regulations and expectations for better transparency by the public.
Clear Standards provides SAP a mature sustainability solution and expertise in carbon management delivered through an agile, Web-based, on-demand delivery model. In order to reduce customer effort and cost around carbon management, SAP will leverage its business process expertise to enable Clear Standards to tap into financial and other data stored in enterprise solutions such as SAP Business Suite 7 software and the SAP Environment, Health, and Safety Management application.
Das further added that SAP looked at social accountability and corporate social responsibility (CSR) as well. These include: economic bottomline, environmental bottomline and social bottomline, respectively.
SAP has 75 percent SMEs as customers and the enterprises make up the remaining 25 percent. Reliance, ONGC, etc., are among its major Indian customers. Das said, "We enable customers to do three things -- help automate, and if automated, optimize; provide companies with best practices -- and Indian companies are certainly very interested in best practices; and enable businesses/users at all levels to make decisions."
He added that ERP is a subset of all processes. In this respect, SAP is much more than an ERP company. Commenting on Business ByDesign, he added that it has been introduced in six countries. "When we are ready, India will be one of the key markets. It will help a lot of customers, especially, the SMEs, to grow," he added.
SAP BusinessObjects Explorer
Sanjay Deshmukh, vice president, Business User and Platforms, SAP India Pvt Ltd, took the opportunity to introduce the SAP BusinessObjects Explorer.Sanjay Deshmukh, vice president, Business User and Platforms, SAP India Pvt Ltd, making a point, while Ranjan Das, managing director, SAP Indian Subcontinent, SAP India Pvt Ltd, looks on.
This tool -- the SAP BusinessObjects Explorer software -- enables you to put the power of business intelligence (BI) into the hands of all of your business users, arming any employee with the clarity and insight to act more quickly and make smarter, more effective business decisions.
According to Deshmukh, a majority of the casual users are underserved by traditional BI tools. A majority of users tend to use search engines while looking for information. The SAP BusinessObjects Explorer software makes use of the power of the Web as well!
This solution provides intuitive experience, data and speed -- rather, immediate insight into mountains of data! "Explorer is a self service, search driven and discovery solution for exploring and analyzing your corporate data," added Deshmukh. "It brings BI to all business users and helps IT to become successful." Among SAP APJ's ramp-up customers are names such as Mahindra and Innogence.
He estimated the Indian BI market to be worth $75-100 million, and about 25-30 percent market share -- the highest market share as per analysts.
Deshmukh added that as a standalone entity, BusinessObjects had the largest share in BI. "Since being bought by SAP, we have also provided clear, futuristic view landscape for customers. Second, it helped us understand business processes better," he said. "Third, SAP also brought in a lot of solutions, such as EPMS, GRC, etc. BI is just a stepping stone toward becoming a clear enterprise."
Before I end, may I take this opportunity to thank SAP for inviting me to this wonderful conference!
Wednesday, July 22, 2009
Zebu-Server -- Enterprise-type emulator from EVE
Friends, this is a guest post by Ms Usha Prasad, Associate Partner, PC Mediaworks. May I take this opportunity to welcome you to my blog, Usha!
Recently, EVE had launched the ZeBu-Server. Let us find out more about it!
EVE Design Automation Ltd., a leader in hardware/software co-verification, has ushered in a new era of hardware-assisted verification with the launch of the ZeBu-Server, a scalable emulation system capable of handling up to one-billion ASIC gates.
The ZeBu-Server is housed in a compact chassis with a small footprint and sets a new standard for small size, light weight and low power consumption making it an environment friendly emulator.
Addressing needs of current, future ASIC/SoC verification
EVE launched ZeBu-Server at a time when the entire industry is quite silent about launching new products. Driven by market demand, EVE, with its strong R&D capabilities rolled out the ZeBu-Server to address the needs of current and future ASIC/SoC verification.
With designs getting larger, growth in embedded software content, verification sequences getting longer and more complex, there is a need for emulation products with higher capacity, higher performance and smarter debugging features. Covering all of these aspects, EVE has chalked out a clear cut R&D roadmap with the ZeBu-Server's launch.
This particular solution is EVE’s next generation emulation box, based on V-5 LX330 FPGAs, with a total design capacity of one billion ASIC-equivalent gates. It is scalable from 10 million gates in increments of 10-, 20-, and 40-million gates.
Providing details on the ZeBu-Server, Montu Makadia, EVE’s director and country sales manager, India, said: "It is a completely new emulator with many new features and capabilities, such as multi-user, relocation capability and multi-RTB (reconfigurable test bench) that increases the data transfer rate in transaction mode to 5 million transactions per second."
Multi-core capabilities
Zebu-Server has multi-core capabilities as well! "A brand new compiler based on multi-core technology has been specifically architected for large, multi-core designs, and it provides fast incremental and parallel compilation. Zebu-Server can map designs of 50+ million gates at a rate of 30 million gates per hour on a farm of PCs. But very large (hundreds of millions of gates) multi-core designs can be mapped at 100 million gates per hours," he added.
"The multi-core feature breaks the design compilation into multiple sub-blocks, and enables the process to be parallelized, and run on a PC farm. It supports module-based scalability. The system is comprised of multiple units and each unit can be configured with up to five modules for a total capacity of 200M ASIC gates. A fully configured system includes five units (25 modules)," he explained.
Smart debugging capabilities
The other important highlight of the ZeBu-Server is its smart debugging capabilities such as SystemVerilog Assertion support and run-time access to every register and signal in the design.
It can map designs of 50+ million gates at a rate of 30 million gates per hour on a farm of PCs. However, very large multi-core designs can be mapped at 100 million gates per hours, as mentioned earlier.
Potential users of ZeBu-Server
Naturally, it would be apt to determine who would be the potential users of ZeBu-Server.
Not a surprise really, that actually, that eight out of the top 10 semiconductor companies are using ZeBu. "We not only see newer applications on processors and graphic chips to be emulated, but also see our existing clients scaling up on their designs. These are the key benefits for them while looking to migration with better speed performance that no traditional emulator can give with such larger design capacity that the ZeBu-Server can handle," Makadia said.
New users will be benefited from ZeBu-Server's exceptional capacity with the advantage of ZeBu's high performance. "ZeBu is superior to that of any traditional emulator," he contended.
Powerful design debugging
ZeBu-Server also helps in powerful design debugging. According to Makadia, ZeBu follows the FPGA flow. Having a very strong software is a key patent EVE enjoys, thus offering very smooth and powerful debugging features. "As per our experience, along with speed-through transactors, it is equally important to have good debugging."
Elaborating further, Makadia said: "Our software helps clients to debug their designs with key features like at-speed static probes, trace memory and logic analyzer triggers to capture signals. However, beyond this, the ZeBu-Server also provides pre-compiled, high-speed flexible probes, which can be used to trace signals directly on to a hard disk, thus giving the user a virtually unlimited trace window.
"Support for SystemVerilog assertions makes it easier to isolate bugs during billions of cycles of emulation. Our enhanced dynamic probes provide the user with run-time access to any register, memory or signal in the design, without recompilation, which enables detailed design debug."
Results a combo of test labs and actual users
It would be interesting to know whether the results are from the test labs or actual users.
The features in Zebu-Server has been embedded with results from test labs and actual users. According to Makadia, it is an enhancement of the features of EVE’s existing XXL box and software. XXL is very stable and used widely by most of EVE's clients.
EVE has tested equally larger and complex designs that will run much faster on Server additionally, with its new Z-Fast Synthesis tool locked with Zebu compiler flow and faster Flexi probes to debug designs while verifying. "These features will surely help customers in a much better and faster way for larger designs," he said.
Current ZeBu customers are already using EVE's fast synthesis tool, zFAST, which is shown to perform 10x faster than third-party synthesis tools, and greatly improving the compilation time. The debug technology is already in use with the XXL, and has been verified in production.
Future enhancements on EVE’s roadmap include very fast design loading of few seconds regardless of the design size.
Key features of Zebu-Server
* Lowest cost-of-ownership.
* Leveraging the largest design capacity.
* Fastest compilation speed on designs exceeding 50 million ASIC-equivalent gates.
* Fastest execution speed.
* Multiple concurrent users.
* Most powerful design debugging.
* Most cycles per dollar in the industry.
Monday, July 20, 2009
Global semicon mid-year review: Not a blip, but recovery won't be smooth, says iSuppli
According to a release from iSuppli today, following four consecutive quarters of reductions, global inventories of chips have declined to appropriate levels, clearing the way for stockpile rebuilding and higher sales in the second half of the year.
However, this particular blog post looks at a previous study from iSuppli where it trimmed the 2009 chip and electronics forecasts, but sees second-half rebound.
Thanks to my good friends, Jon Cassell and Debra Jaramilla, I was able to connect with Dale Ford, senior vice president, market intelligence services, for iSuppli, for a discussion on this particular study.
Given the lingering economic woes and continuing poor visibility into future demand trends, iSuppli, as mentioned, reduced its forecasts for global semiconductor and electronic equipment revenue in 2009.
Worldwide electronic equipment revenue is set to decline to $1.38 trillion in 2009, down 9.8 percent from $1.53 trillion in 2008. iSuppli's previous forecast in April predicted a 7.6 percent decline in revenue. Global semiconductor revenue is set to fall to $198.9 billion in 2009, down 23 percent from $258.5 billion in 2008. iSuppli's April forecast called for a 21.5 percent decline.
I quizzed Ford on how iSuppli sees the global semicon market performing over the rest of the year. He said: "We do expect sequential improvement in the second half of 2009 compared to the first half of 2009. However, the market will remain significantly below the market level of the same period in 2008."
It is also apt to determine the behaviour of the electronic equipment segment in the same period. iSuppli expects a similar pattern in the electronics segment based on the current economic outlook and guidance from key OEMs.
According to iSuppli's study, while all of the electronics segments are likely to suffer contractions in 2009, the automotive sector was a major culprit behind the downgrade. So, is it fair to only blame the automotive sector for the downgrade?
No, said Ford. "It should be noted that the downgrade is very minor and reflects a more broad-based impact of the economy on the electronics market."
I mentioned about a new study from iSuppli, which talks about global chip inventories declining to appropriate levels, clearing the way for stockpile rebuilding and higher sales during H2-2009.
Prior to the release of this study, I had asked Dale Ford and iSuppli whether we would see companies revising their forecasts.
According to Ford, companies are already revising their forecasts and this has been noted in some of the earnings announcements this week. However, there is still great uncertainty in the economy and this presents the likelihood that company and industry expectations will continue to fluctuate.
As for iSuppli's growth prediction for H2-09, its current published second half outlook calls for H2-09 to grow by over 17 percent compared to H1-09. iSuppli also is sticking to the 13.1 percent growth for the semiconductor industry in 2010.
As for the factors now leading to conditions looking up in H2, Ford said that the global economy is the dominant factor driving industry growth. All other factors are secondary in comparison.
It would also be interesting to see how the Japanese semiconductor industry is likely to hold out in H2. Ford added: "Japanese semiconductor suppliers have experienced a very difficult H1. However, with improving consumer sentiment, they have the opportunity for an improved H2."
There is also a clear indication of the starting of the correction phase to rebalance over-depleted inventories. Ford said that a number of companies have noted that their Q2 sales are influenced both by rebalancing inventories and expectations of H2-09 demand.
If firms still continue in a wait-and-see mode, won't this serve to make the market dynamics worse as firms are then faced with a massive catch-up problem? What should they do?
Ford advised: "There is the potential for companies to miss opportunities in the market if they are too passive. Companies need to aggressively communicate with their partners and clients to obtain the best visibility possible in planning their tactics and strategies for meeting market demand."
Recovery or blip?
Finally, is this really the start of the chip market recovery or a blip on the statistics radar screen?
He said: "While we do not consider this to be a 'blip', we do expect that the 'recovery' will not be smooth. Ongoing uncertainty and volatility in the economy will impact the progress of the market moving forward."
However, this particular blog post looks at a previous study from iSuppli where it trimmed the 2009 chip and electronics forecasts, but sees second-half rebound.
Thanks to my good friends, Jon Cassell and Debra Jaramilla, I was able to connect with Dale Ford, senior vice president, market intelligence services, for iSuppli, for a discussion on this particular study.
Given the lingering economic woes and continuing poor visibility into future demand trends, iSuppli, as mentioned, reduced its forecasts for global semiconductor and electronic equipment revenue in 2009.
Worldwide electronic equipment revenue is set to decline to $1.38 trillion in 2009, down 9.8 percent from $1.53 trillion in 2008. iSuppli's previous forecast in April predicted a 7.6 percent decline in revenue. Global semiconductor revenue is set to fall to $198.9 billion in 2009, down 23 percent from $258.5 billion in 2008. iSuppli's April forecast called for a 21.5 percent decline.
I quizzed Ford on how iSuppli sees the global semicon market performing over the rest of the year. He said: "We do expect sequential improvement in the second half of 2009 compared to the first half of 2009. However, the market will remain significantly below the market level of the same period in 2008."
It is also apt to determine the behaviour of the electronic equipment segment in the same period. iSuppli expects a similar pattern in the electronics segment based on the current economic outlook and guidance from key OEMs.
According to iSuppli's study, while all of the electronics segments are likely to suffer contractions in 2009, the automotive sector was a major culprit behind the downgrade. So, is it fair to only blame the automotive sector for the downgrade?
No, said Ford. "It should be noted that the downgrade is very minor and reflects a more broad-based impact of the economy on the electronics market."
I mentioned about a new study from iSuppli, which talks about global chip inventories declining to appropriate levels, clearing the way for stockpile rebuilding and higher sales during H2-2009.
Prior to the release of this study, I had asked Dale Ford and iSuppli whether we would see companies revising their forecasts.
According to Ford, companies are already revising their forecasts and this has been noted in some of the earnings announcements this week. However, there is still great uncertainty in the economy and this presents the likelihood that company and industry expectations will continue to fluctuate.
As for iSuppli's growth prediction for H2-09, its current published second half outlook calls for H2-09 to grow by over 17 percent compared to H1-09. iSuppli also is sticking to the 13.1 percent growth for the semiconductor industry in 2010.
As for the factors now leading to conditions looking up in H2, Ford said that the global economy is the dominant factor driving industry growth. All other factors are secondary in comparison.
It would also be interesting to see how the Japanese semiconductor industry is likely to hold out in H2. Ford added: "Japanese semiconductor suppliers have experienced a very difficult H1. However, with improving consumer sentiment, they have the opportunity for an improved H2."
There is also a clear indication of the starting of the correction phase to rebalance over-depleted inventories. Ford said that a number of companies have noted that their Q2 sales are influenced both by rebalancing inventories and expectations of H2-09 demand.
If firms still continue in a wait-and-see mode, won't this serve to make the market dynamics worse as firms are then faced with a massive catch-up problem? What should they do?
Ford advised: "There is the potential for companies to miss opportunities in the market if they are too passive. Companies need to aggressively communicate with their partners and clients to obtain the best visibility possible in planning their tactics and strategies for meeting market demand."
Recovery or blip?
Finally, is this really the start of the chip market recovery or a blip on the statistics radar screen?
He said: "While we do not consider this to be a 'blip', we do expect that the 'recovery' will not be smooth. Ongoing uncertainty and volatility in the economy will impact the progress of the market moving forward."
Sunday, July 19, 2009
Global semicon mid-year review: Chip market revival or blip on stats radar screen?
A recent report from Future Horizons suggests an 18 percent growth for the chip market in Q2-2009! So, is this a sign of the chip market recovery or a mere blip on the statistics radar screen?
It is both, said, Malcolm Penn, chairman, founder and CEO of Future Horizons, and counselled that: "The fourth quarter market collapse was far too steep -- a severe over-reaction to last year's gross financial uncertainty -- culminating with the Lehman Brothers collapse in September. The first quarter saw this stabilise with the second quarter restocking, but there are other positive factors also in play."
Examining a bit further, here's what he further revealed. One, the memory market is seeing some signs of slow recovery. He said, "This has already started DDR3 driven!" Likewise, companies are also in the process of revising their forecasts. The reason, Penn contended, being, "The maths has changed dramatically since Jan 2009!"
According to him, factors now leading to conditions looking up in H2 2009, include the normal seasonal demand -- from a tight inventory base -- and tightening capacity. There is also a clear indication of the correction phase to rebalance over-depleted inventories having started. "This is what's driving Q2's high unit, and therefore, sales growth," he contended.
Firms advised to stop seeing and waiting!
This isn't all! Penn further counselled firms who are still in a wait-and-see mode to 'stop seeing and waiting'! Next, fabs are also looking to maximize their returns. For one, they have stopped over-investing.
Do we have enough stats from others to back up what's been happening in the global semiconductor industry? Perhaps, yes!
IC Insights stands out
First, look at IC Insights! It has stood out by pointing out in early July that H2-09 is likely to usher in strong seasonal strength for electronic system sales, a period of IC inventory replenishment, which began in 2Q09, and positive worldwide GDP growth.
IC Insights has predicted global IC market to grow +18 percent; IC foundry sales to grow +43 percent; and semiconductor capital spending to grow +28 percent in H2-09.
DDR3 driving memory recovery? Flat NAND?
Elsewhere, Converge Market Insights said that according to major DRAM manufacturers, DDR3 demand has been on the rise over the last two months and supply is limited.
This is quite in line with Future Horizons contention that there is a DDR3 driven memory recovery, albeit slow. It would be interesting to see how Q3-09 plays out.
As for NAND, according to DRAMeXchange, the NAND market may continue to show the tug-of-war status in July due to dissimilar positive and negative market factors perceived and expected by both sides. As a result, NAND Flash contract prices are likely to somewhat soften or stay flat in the short term.
Semicon equipment market to decline 52 percent in 2009!
According to SEMI, it projects 2009 semiconductor equipment sales to reach $14.14 billion as per the mid-year edition of the SEMI Capital Equipment Forecast, released by SEMI at the annual SEMICON West exposition.
The forecast indicates that, following a 31 percent market decline in 2008, the equipment market will decline another 52 percent in 2009, but will experience a rebound with annual growth of about 47 percent in 2010.
EDA cause for concern
The EDA industry still remains a cause for concern. The EDA Consortium's Market Statistics Service (MSS) announced that the EDA industry revenue for Q1 2009 declined 10.7 percent to $1,192.1 million, compared to $1,334.2 million in Q1 2008, driven primarily by an accounting shift at one major EDA company. The four-quarter moving average declined 11.3 percent.
If you look at the last five quarters, the EDA industry has really been having it rough. Here are the numbers over the last five quarters, as per the Consortium:
* The EDA industry revenue for Q1 2008 declined 1.2 percent to $1,350.7 million compared to $1,366.8 million in Q1 2007.
* The industry revenue for Q2 2008 declined 3.7 percent to $1,357.4 million compared to $1,408.8 million in Q2 2007.
* The industry revenue for Q3 2008 declined 10.9 percent to $1,258.6 million compared to $1,412.1 million in Q3 2007.
* The industry revenue for Q4 2008 declined 17.7 percent to $1,318.7 million, compared to $1,602.7 million in Q4 2007.
Therefore, at the end of the day, what do you have? For now, the early recovery signs are more of a blip on the stats radar screen and there's still some way to go and work to be done before the global semiconductor industry can clearly proclaim full recovery!
Before I close, a word about the Indian semiconductor industry. Perhaps, it needs to start moving a bit faster and quicker than it is doing presently. Borrowing a line from Malcolm Penn, the Indian semiconductor industry surely needs to "stop waiting and watching."
I will be in conversation next with iSuppli on the chip and electronics industry forecasts. Keep watching this space, friends.
It is both, said, Malcolm Penn, chairman, founder and CEO of Future Horizons, and counselled that: "The fourth quarter market collapse was far too steep -- a severe over-reaction to last year's gross financial uncertainty -- culminating with the Lehman Brothers collapse in September. The first quarter saw this stabilise with the second quarter restocking, but there are other positive factors also in play."
Examining a bit further, here's what he further revealed. One, the memory market is seeing some signs of slow recovery. He said, "This has already started DDR3 driven!" Likewise, companies are also in the process of revising their forecasts. The reason, Penn contended, being, "The maths has changed dramatically since Jan 2009!"
According to him, factors now leading to conditions looking up in H2 2009, include the normal seasonal demand -- from a tight inventory base -- and tightening capacity. There is also a clear indication of the correction phase to rebalance over-depleted inventories having started. "This is what's driving Q2's high unit, and therefore, sales growth," he contended.
Firms advised to stop seeing and waiting!
This isn't all! Penn further counselled firms who are still in a wait-and-see mode to 'stop seeing and waiting'! Next, fabs are also looking to maximize their returns. For one, they have stopped over-investing.
Do we have enough stats from others to back up what's been happening in the global semiconductor industry? Perhaps, yes!
IC Insights stands out
First, look at IC Insights! It has stood out by pointing out in early July that H2-09 is likely to usher in strong seasonal strength for electronic system sales, a period of IC inventory replenishment, which began in 2Q09, and positive worldwide GDP growth.
IC Insights has predicted global IC market to grow +18 percent; IC foundry sales to grow +43 percent; and semiconductor capital spending to grow +28 percent in H2-09.
DDR3 driving memory recovery? Flat NAND?
Elsewhere, Converge Market Insights said that according to major DRAM manufacturers, DDR3 demand has been on the rise over the last two months and supply is limited.
This is quite in line with Future Horizons contention that there is a DDR3 driven memory recovery, albeit slow. It would be interesting to see how Q3-09 plays out.
As for NAND, according to DRAMeXchange, the NAND market may continue to show the tug-of-war status in July due to dissimilar positive and negative market factors perceived and expected by both sides. As a result, NAND Flash contract prices are likely to somewhat soften or stay flat in the short term.
Semicon equipment market to decline 52 percent in 2009!
According to SEMI, it projects 2009 semiconductor equipment sales to reach $14.14 billion as per the mid-year edition of the SEMI Capital Equipment Forecast, released by SEMI at the annual SEMICON West exposition.
The forecast indicates that, following a 31 percent market decline in 2008, the equipment market will decline another 52 percent in 2009, but will experience a rebound with annual growth of about 47 percent in 2010.
EDA cause for concern
The EDA industry still remains a cause for concern. The EDA Consortium's Market Statistics Service (MSS) announced that the EDA industry revenue for Q1 2009 declined 10.7 percent to $1,192.1 million, compared to $1,334.2 million in Q1 2008, driven primarily by an accounting shift at one major EDA company. The four-quarter moving average declined 11.3 percent.
If you look at the last five quarters, the EDA industry has really been having it rough. Here are the numbers over the last five quarters, as per the Consortium:
* The EDA industry revenue for Q1 2008 declined 1.2 percent to $1,350.7 million compared to $1,366.8 million in Q1 2007.
* The industry revenue for Q2 2008 declined 3.7 percent to $1,357.4 million compared to $1,408.8 million in Q2 2007.
* The industry revenue for Q3 2008 declined 10.9 percent to $1,258.6 million compared to $1,412.1 million in Q3 2007.
* The industry revenue for Q4 2008 declined 17.7 percent to $1,318.7 million, compared to $1,602.7 million in Q4 2007.
Therefore, at the end of the day, what do you have? For now, the early recovery signs are more of a blip on the stats radar screen and there's still some way to go and work to be done before the global semiconductor industry can clearly proclaim full recovery!
Before I close, a word about the Indian semiconductor industry. Perhaps, it needs to start moving a bit faster and quicker than it is doing presently. Borrowing a line from Malcolm Penn, the Indian semiconductor industry surely needs to "stop waiting and watching."
I will be in conversation next with iSuppli on the chip and electronics industry forecasts. Keep watching this space, friends.
Wednesday, July 15, 2009
Cadence integrates chip planning with implementation!
Last month, Cadence Design Systems Inc. unveiled an integrated chip planning and implementation solution. This has been achieved through the integration of Cadence InCyte Chip Estimator and the Cadence Encounter Digital Implementation (EDI) System technologies.
Cadence has called this breakthrough solution, which provides design and implementation engineers with superior visibility and predictability of chip performance, area, power consumption, cost, and time to market across the full range of design activities, including system-level design and IP selection through final implementation and signoff.
I got into a brief conversation with Adam Traidman, Group Marketing Director, Cadence, Dave Desharnais, Product Marketing Group Director, Cadence, and Rahul Arya, Director, Marketing & Technology Sales, Cadence Design Systems India Pvt Ltd.
EDA industry revenue dips 10.7 percent in Q1-09
By the way, the EDA Consortium (EDAC) Market Statistics Service (MSS) today announced that the EDA industry revenue for Q1 2009 declined 10.7 percent to $1192.1 million, compared to $1334.2 million in Q1 2008, driven primarily by an accounting shift at one major EDA company. The four-quarter moving average declined 11.3 percent.
"The business environment remained difficult for EDA as for other industries, with Q1 EDA revenues declining in all regions except Asia Pacific," said Walden C. Rhines, EDA Consortium chair and Mentor Graphics CEO and chairman. "Nevertheless, for Q1, the overall decline was less than for the previous quarter."
Back to the current discussion then! It'd be interesting to see how all these tools bring the EDA industry back above the red level!
Why this solution?
The obvious question, why the Integrated Chip Planning and Implementation Solution now?
Adam Traidman said that the Chip Estimator is quite unique! It helps customers early in the IC design cycle.
"We go beyond EDA and estimate cost, etc. We help the designers to do an early architectural level ecomnomical and techical analysis and estimation, etc. Statistics show that during the early phases of design, those decisions can contribute to 80 percent of final design. Today, very few EDA companies provide set of tools and methodologies that allow such trade-off," he added.
According to him, every customer does this analysis, probably, manually. Cadence is now automating this method. In this respect, it has integrated chip planning with implementation.
"The results of the analysis -- you are concerned about accuracy; you look to the EDA vendors to help converge from initial implementation to the actual convergence. Think of it like a cockpit for the design engineer, general manager, program manager, etc.," he noted.
"You've made all the fundamental decisions, etc. If you're sitting on the physical implementation tool, and you need think through the implications that can be there. For example, to re-synthesize new libraries, etc. We are talking about chip planning at a much, much higher level," added Traidman.
Helping with IP selection!
The Cadence solution also leverages the vast ecosystem of IP at the ChipEstimate.com portal where over 200 IP suppliers and foundries contribute data. Helping with IP selection has been mentioned among the processes, perhaps, an indicator that designers may have not been able to select the right IPs all this while.
According to Traidman, IP selection and qiuality are key issues. "A lot of people, doing these tradeoffs, could be design managers, general manager, etc. When they sit with this tool, and when it pops up, they can see a huge library of 7,000 IPs from about 200 IP suppliers and foundries. Any design team can view all of the IPs as a free service," he elaborated. By the way, ChipEstimate is owned by Cadence!
He further added that the ChipEstimate portal allows customers to lower the risks of converging. The portal has been growing since 2006, and receives 1 million page view each month.
Just for interest's sake, there's another site -- Design And Reuse -- that claims to be the world's largest directory of 8,000 silicon IPs from more than 400 vendors! I have also got into some other discussions -- that are ongoing -- for developing a similar site in India, for the Indian semiconductor industry!
What about Cadence Encounter?
Post the integration, what happens now to the Cadence Encounter solution and whether it is still available standalone?
Dave Desharnais said the Cadence Encounter solution is still available standalone. "We have integrated some key functions from InCyte. From InCyte, you would normally not have the link to get into physical implementation. Likewise, with feeding back of a fully realized database," he said.
Last December (2008), Cadence had announced the Encounter Digital Implementation System, a next generation complete RTL-to-GDSII solution for logic and physical implementation.
Along with a fundamental new memory architecture and end-to-end multicore backplane to address the requirements of leapfrog capacity and faster turnaround time for billion transistor designs, it also delivers complete implementation and signoff-in-the-loop for low power, mixed signal, and advanced node design; including the latest 28nm process node where it has been used on over half of the designs being done at this node today.
As per Rahul Arya, since the initial launch, there has been significant usage and endorsement from the world's largest semiconductor companies, including ST, Toshiba, NEC, NXP, Fujitsu, AMD, and many more that are requested as non-public endorsements.
He added: "The announcement of InCyte and EDI System integration brings a whole new dimension for both system-level and design implementation teams. While both solutions -- InCyte and EDI System -- are still available as standalone, using both solutions together enables designers at all levels to now have complete visibility into all aspects of the design -- from system level architecture requirements and IP selection, to full physical floorplanning, final low power and implementation signoff results.
"The bringing together of both of these solutions delivers literally unprecedented predictability, visibility, and accuracy into all steps of the chip creation and implementation flow for faster design convergence."
The design solution will be demonstrated at the Design Automation Conference (DAC 2009) in San Francisco this month and made available later this year.
Cadence has called this breakthrough solution, which provides design and implementation engineers with superior visibility and predictability of chip performance, area, power consumption, cost, and time to market across the full range of design activities, including system-level design and IP selection through final implementation and signoff.
I got into a brief conversation with Adam Traidman, Group Marketing Director, Cadence, Dave Desharnais, Product Marketing Group Director, Cadence, and Rahul Arya, Director, Marketing & Technology Sales, Cadence Design Systems India Pvt Ltd.
EDA industry revenue dips 10.7 percent in Q1-09
By the way, the EDA Consortium (EDAC) Market Statistics Service (MSS) today announced that the EDA industry revenue for Q1 2009 declined 10.7 percent to $1192.1 million, compared to $1334.2 million in Q1 2008, driven primarily by an accounting shift at one major EDA company. The four-quarter moving average declined 11.3 percent.
"The business environment remained difficult for EDA as for other industries, with Q1 EDA revenues declining in all regions except Asia Pacific," said Walden C. Rhines, EDA Consortium chair and Mentor Graphics CEO and chairman. "Nevertheless, for Q1, the overall decline was less than for the previous quarter."
Back to the current discussion then! It'd be interesting to see how all these tools bring the EDA industry back above the red level!
Why this solution?
The obvious question, why the Integrated Chip Planning and Implementation Solution now?
Adam Traidman said that the Chip Estimator is quite unique! It helps customers early in the IC design cycle.
"We go beyond EDA and estimate cost, etc. We help the designers to do an early architectural level ecomnomical and techical analysis and estimation, etc. Statistics show that during the early phases of design, those decisions can contribute to 80 percent of final design. Today, very few EDA companies provide set of tools and methodologies that allow such trade-off," he added.
According to him, every customer does this analysis, probably, manually. Cadence is now automating this method. In this respect, it has integrated chip planning with implementation.
"The results of the analysis -- you are concerned about accuracy; you look to the EDA vendors to help converge from initial implementation to the actual convergence. Think of it like a cockpit for the design engineer, general manager, program manager, etc.," he noted.
"You've made all the fundamental decisions, etc. If you're sitting on the physical implementation tool, and you need think through the implications that can be there. For example, to re-synthesize new libraries, etc. We are talking about chip planning at a much, much higher level," added Traidman.
Helping with IP selection!
The Cadence solution also leverages the vast ecosystem of IP at the ChipEstimate.com portal where over 200 IP suppliers and foundries contribute data. Helping with IP selection has been mentioned among the processes, perhaps, an indicator that designers may have not been able to select the right IPs all this while.
According to Traidman, IP selection and qiuality are key issues. "A lot of people, doing these tradeoffs, could be design managers, general manager, etc. When they sit with this tool, and when it pops up, they can see a huge library of 7,000 IPs from about 200 IP suppliers and foundries. Any design team can view all of the IPs as a free service," he elaborated. By the way, ChipEstimate is owned by Cadence!
He further added that the ChipEstimate portal allows customers to lower the risks of converging. The portal has been growing since 2006, and receives 1 million page view each month.
Just for interest's sake, there's another site -- Design And Reuse -- that claims to be the world's largest directory of 8,000 silicon IPs from more than 400 vendors! I have also got into some other discussions -- that are ongoing -- for developing a similar site in India, for the Indian semiconductor industry!
What about Cadence Encounter?
Post the integration, what happens now to the Cadence Encounter solution and whether it is still available standalone?
Dave Desharnais said the Cadence Encounter solution is still available standalone. "We have integrated some key functions from InCyte. From InCyte, you would normally not have the link to get into physical implementation. Likewise, with feeding back of a fully realized database," he said.
Last December (2008), Cadence had announced the Encounter Digital Implementation System, a next generation complete RTL-to-GDSII solution for logic and physical implementation.
Along with a fundamental new memory architecture and end-to-end multicore backplane to address the requirements of leapfrog capacity and faster turnaround time for billion transistor designs, it also delivers complete implementation and signoff-in-the-loop for low power, mixed signal, and advanced node design; including the latest 28nm process node where it has been used on over half of the designs being done at this node today.
As per Rahul Arya, since the initial launch, there has been significant usage and endorsement from the world's largest semiconductor companies, including ST, Toshiba, NEC, NXP, Fujitsu, AMD, and many more that are requested as non-public endorsements.
He added: "The announcement of InCyte and EDI System integration brings a whole new dimension for both system-level and design implementation teams. While both solutions -- InCyte and EDI System -- are still available as standalone, using both solutions together enables designers at all levels to now have complete visibility into all aspects of the design -- from system level architecture requirements and IP selection, to full physical floorplanning, final low power and implementation signoff results.
"The bringing together of both of these solutions delivers literally unprecedented predictability, visibility, and accuracy into all steps of the chip creation and implementation flow for faster design convergence."
The design solution will be demonstrated at the Design Automation Conference (DAC 2009) in San Francisco this month and made available later this year.
Monday, July 13, 2009
Q2-09 chip market shows 18 percent growth: Mid-year semicon update
Here are the excerpts from the Global Semiconductor Monthly Report, July 2009, provided by Malcolm Penn, chairman, founder and CEO of Future Horizons.
Following hot on the heels of April's 16 percent month-on-month sales growth, May grew a further 0.9 percent sequentially, putting June on track to break through the $20 billion barrier, for the first time since the chip market collapsed last September.
"Psychologically this will give everyone a shot in the arm," commented Malcolm Penn. "Second quarter growth is usually pretty pathetic; there have been only three historical precedents when such a spurt has happened. The big question now is: Is this the start of the chip market recovery or is it merely a blip on the statistics radar screen?"
"It's both," Penn counselled. "The fourth quarter market collapse was far too steep -- a severe over-reaction to last year's gross financial uncertainty -- culminating with the Lehman Brothers collapse in September. The first quarter saw this stabilise with the second quarter restocking, but there are other positive factors also in play."
The normal dynamics following any market collapse is: (1) over-reaction, cutting back production and inventories too far, (2) a correction phase to rebalance over-depleted inventories, and (3) a resumption of demand-driven built.
"Whilst in Phase 1, chip sales equals OEM requirement minus inventory burn, hence understating real demand, Phase 2 results in sales equal to OEM demand plus inventory rebuild, thus overstating the actual demand," continued Penn. "We are currently in Phase 2 of the recovery cycle."
Underlying issues
Commenting on some of the other underlying issues, Future Horizons believes that the timing of this recovery is 'as good as it gets', given that third quarter seasonal demand will be on the increase just as the sales decline caused by inventory build ends.
Capacity will also shortly start to come into play, given the unprecedented three years of fab under-investment. Traditionally tight availability positively impacts IC ASPs, but delayed by 12 months as existing contracts run their course.
This time, Future Horizons believes things could be faster, given the extent of the cutbacks, as already witnessed in the 2009 memory market.
"This (under-investment) eagle is one day coming home to roost," warned Penn. "We are already seeing the first signs of shortages at UMC affecting Xilinx and other firms second quarter sales. And for shortages read ASP increases, it's inevitable, it's just a matter of time, the fabs will be looking to maximise the return on their costly and precious resources, especially now more and more firms are seeking their share of the foundry pie. Better to pay 5x the price and get the parts you need."
On the cautionary side, the overall economic outlook remains uncertain and moribund, and firms are still in a wait-and-see mode, which ironically will only serve to make the market dynamics worse, as firms are then faced with a massive catch-up problem.
"Never forget market trends are not based just on rational decisions, but emotional ones as well, it is this that makes the outcome sometimes so difficult to predict," Penn reminded.
I hope to get into further conversation with Malcolm Penn on this recovery or blip!
Following hot on the heels of April's 16 percent month-on-month sales growth, May grew a further 0.9 percent sequentially, putting June on track to break through the $20 billion barrier, for the first time since the chip market collapsed last September.
"Psychologically this will give everyone a shot in the arm," commented Malcolm Penn. "Second quarter growth is usually pretty pathetic; there have been only three historical precedents when such a spurt has happened. The big question now is: Is this the start of the chip market recovery or is it merely a blip on the statistics radar screen?"
"It's both," Penn counselled. "The fourth quarter market collapse was far too steep -- a severe over-reaction to last year's gross financial uncertainty -- culminating with the Lehman Brothers collapse in September. The first quarter saw this stabilise with the second quarter restocking, but there are other positive factors also in play."
The normal dynamics following any market collapse is: (1) over-reaction, cutting back production and inventories too far, (2) a correction phase to rebalance over-depleted inventories, and (3) a resumption of demand-driven built.
"Whilst in Phase 1, chip sales equals OEM requirement minus inventory burn, hence understating real demand, Phase 2 results in sales equal to OEM demand plus inventory rebuild, thus overstating the actual demand," continued Penn. "We are currently in Phase 2 of the recovery cycle."
Underlying issues
Commenting on some of the other underlying issues, Future Horizons believes that the timing of this recovery is 'as good as it gets', given that third quarter seasonal demand will be on the increase just as the sales decline caused by inventory build ends.
Capacity will also shortly start to come into play, given the unprecedented three years of fab under-investment. Traditionally tight availability positively impacts IC ASPs, but delayed by 12 months as existing contracts run their course.
This time, Future Horizons believes things could be faster, given the extent of the cutbacks, as already witnessed in the 2009 memory market.
"This (under-investment) eagle is one day coming home to roost," warned Penn. "We are already seeing the first signs of shortages at UMC affecting Xilinx and other firms second quarter sales. And for shortages read ASP increases, it's inevitable, it's just a matter of time, the fabs will be looking to maximise the return on their costly and precious resources, especially now more and more firms are seeking their share of the foundry pie. Better to pay 5x the price and get the parts you need."
On the cautionary side, the overall economic outlook remains uncertain and moribund, and firms are still in a wait-and-see mode, which ironically will only serve to make the market dynamics worse, as firms are then faced with a massive catch-up problem.
"Never forget market trends are not based just on rational decisions, but emotional ones as well, it is this that makes the outcome sometimes so difficult to predict," Penn reminded.
I hope to get into further conversation with Malcolm Penn on this recovery or blip!
Sunday, July 12, 2009
Clearly, mixed signals in OEM semiconductor design activities!
Friends, here is the full report on iSuppli's recent activity titled "Mixed Signals in OEM Design Activities".
Min-Sun Moon, senior analyst, Semiconductor Spend and Design, iSuppli, discussed how the "values" of design activities are discerned globally and how design decisions are made by a given country.
This report should be of particular interest to the Indian semiconductor design industry as it is apparent there is considerable scope for growth and development.
It is very well documented that everyone has been hit hard by the economic downturn. The electronic OEMs are no exception. They have also reduced shipments. The average selling prices (ASPs) of semiconductor devices have dropped dramatically as well.
Top six design influencing countries
As per iSuppli's Design Activity Tool, the top six countries leading in the design influence are as follow: USA, Japan, China/Hong Kong, Taiwan, South Korea and Germany. The United States retains the no. 1 position, followed by Japan and China.
The dramatic changes in ASPs of chips and products meant an almost about 5 percent drop in semiconductor spend in 2008, and above 21 percent drop in semiconductor spend by the top OEMs in 2009. Hence, design activities by top OEMs dropped significantly.
The USA apparently has been going through a tough period, and it does not seem to have a bright future in 2009-10 due to drop in design spends. However, in 2010, it should post about 9-10 percent growth. The top design influencers in the USA include HP, Dell, Apple and Motorola.
China seizes opportunity
According to Moon, Japan retained the second position. However, China has seized the opportunity during the recession. It has some growth compared to other countries who have had negative growth this year.
China still remains one of the most attractive markets for OEMs to enter. Many top OEM have either opened or expanded R&D centers in China in the last few years.
However, because of the recession, the expansion by OEMs slowed down in China during 2009. Nevertheless, the Chinese market continues to grow. In the next few years, China will grow and the other countries will have some positive growth as well, but their growth will be slower than that of China.
China has also been showing interesting signs. Some Chinese companies are trying to enter new markets, such as automotive.
China is currently the third largest country in terms of design influence. The design share is about 10 percent in 2009. China could get close to Japan and the USA, but it will not happen in the near future though.
Top five countries in 2009
In 2009, the top five countries by design influence spend share are as follows: USA -- 31 percent, Japan -- 25 percent, China/Hong Kong -- 10 percent, Taiwan -- 8 percent, South Korea -- 7 percent, and the Rest of the World -- 19 percent.
Mixed signals are apparent in the design activities by country. For instance, this year, the USA has been losing market share. A large percent of design activities are moving to the Asia Pacific region. Some business in the USA is being continued or reduced -- and being moved to other regions -- in order to maintain the business and lower the cost of operations.
Japan's design spend share increased from 22 percent (approximately $40 bn in 2008) to 25 percent in 2009. Japan is bringing a lot of design activities back home.
Taiwan used to be third largest in the design influence, but has now dropped to the fourth position, with share in design spend reaching 8 percent in 2009. China also contributed to the changes here. However, it is still better than others as some OEMs are still outsourcing to some ODMs located in Taiwan.
Identifying targets by regions
iSuppli gave examples of designing with sensors and actuators, and LEDs, as these are very popular currently.
According to Moon, designs using sensors and actuators have been more than 30 percent in the USA, while Japan has more than 25 percent. It is over 20 percent in Europe, while such designs have been less in Asia Pacific -- above 15 percent.
The biggest influencers for sensors and actuators in the USA are said to be Apple, HP and TRW Automotive.
For LEDs, more use has been happening in Japan -- over 30 percent. As an example, there are more LED TV design activities in Japan. The biggest influencers for LEDs in Japan are Canon and Sony.
Changes due to M&A
Another trend visible in the design spend share has been the changes due to mergers and acquisitions.
As an example, we have the Mitac Group, which acquired Magellan's consumer products division. In 2008, Mitac Group had 78 percent spend in Taiwan, and 18 percent in the USA. After acquiring Magellan, Taiwan's design spend share became 57 percent and USA's became 13 percent. On the other hand, France's share grew to 17 percent and Russia's to 7 percent. This indicates that country-wise, budgets do get changed. This is just one example.
These are indeed very interesting numbers and facts, and as mentioned earlier, India has a considerable opportunity as an influencer in the semiconductor design spend going forward.
Min-Sun Moon, senior analyst, Semiconductor Spend and Design, iSuppli, discussed how the "values" of design activities are discerned globally and how design decisions are made by a given country.
This report should be of particular interest to the Indian semiconductor design industry as it is apparent there is considerable scope for growth and development.
It is very well documented that everyone has been hit hard by the economic downturn. The electronic OEMs are no exception. They have also reduced shipments. The average selling prices (ASPs) of semiconductor devices have dropped dramatically as well.
Top six design influencing countries
As per iSuppli's Design Activity Tool, the top six countries leading in the design influence are as follow: USA, Japan, China/Hong Kong, Taiwan, South Korea and Germany. The United States retains the no. 1 position, followed by Japan and China.
The dramatic changes in ASPs of chips and products meant an almost about 5 percent drop in semiconductor spend in 2008, and above 21 percent drop in semiconductor spend by the top OEMs in 2009. Hence, design activities by top OEMs dropped significantly.
The USA apparently has been going through a tough period, and it does not seem to have a bright future in 2009-10 due to drop in design spends. However, in 2010, it should post about 9-10 percent growth. The top design influencers in the USA include HP, Dell, Apple and Motorola.
China seizes opportunity
According to Moon, Japan retained the second position. However, China has seized the opportunity during the recession. It has some growth compared to other countries who have had negative growth this year.
China still remains one of the most attractive markets for OEMs to enter. Many top OEM have either opened or expanded R&D centers in China in the last few years.
However, because of the recession, the expansion by OEMs slowed down in China during 2009. Nevertheless, the Chinese market continues to grow. In the next few years, China will grow and the other countries will have some positive growth as well, but their growth will be slower than that of China.
China has also been showing interesting signs. Some Chinese companies are trying to enter new markets, such as automotive.
China is currently the third largest country in terms of design influence. The design share is about 10 percent in 2009. China could get close to Japan and the USA, but it will not happen in the near future though.
Top five countries in 2009
In 2009, the top five countries by design influence spend share are as follows: USA -- 31 percent, Japan -- 25 percent, China/Hong Kong -- 10 percent, Taiwan -- 8 percent, South Korea -- 7 percent, and the Rest of the World -- 19 percent.
Mixed signals are apparent in the design activities by country. For instance, this year, the USA has been losing market share. A large percent of design activities are moving to the Asia Pacific region. Some business in the USA is being continued or reduced -- and being moved to other regions -- in order to maintain the business and lower the cost of operations.
Japan's design spend share increased from 22 percent (approximately $40 bn in 2008) to 25 percent in 2009. Japan is bringing a lot of design activities back home.
Taiwan used to be third largest in the design influence, but has now dropped to the fourth position, with share in design spend reaching 8 percent in 2009. China also contributed to the changes here. However, it is still better than others as some OEMs are still outsourcing to some ODMs located in Taiwan.
Identifying targets by regions
iSuppli gave examples of designing with sensors and actuators, and LEDs, as these are very popular currently.
According to Moon, designs using sensors and actuators have been more than 30 percent in the USA, while Japan has more than 25 percent. It is over 20 percent in Europe, while such designs have been less in Asia Pacific -- above 15 percent.
The biggest influencers for sensors and actuators in the USA are said to be Apple, HP and TRW Automotive.
For LEDs, more use has been happening in Japan -- over 30 percent. As an example, there are more LED TV design activities in Japan. The biggest influencers for LEDs in Japan are Canon and Sony.
Changes due to M&A
Another trend visible in the design spend share has been the changes due to mergers and acquisitions.
As an example, we have the Mitac Group, which acquired Magellan's consumer products division. In 2008, Mitac Group had 78 percent spend in Taiwan, and 18 percent in the USA. After acquiring Magellan, Taiwan's design spend share became 57 percent and USA's became 13 percent. On the other hand, France's share grew to 17 percent and Russia's to 7 percent. This indicates that country-wise, budgets do get changed. This is just one example.
These are indeed very interesting numbers and facts, and as mentioned earlier, India has a considerable opportunity as an influencer in the semiconductor design spend going forward.
Tuesday, July 7, 2009
Xilinx Base Targeted Design Platform helps save months of development!
Last week, Xilinx discussed its Targeted Design Platforms, aimed at accelerating the development of system-on-chip (SoC) solutions with Xilinx Virtex-6 and Spartan-6 FPGAs.
I was in conversation with Brent Przybus, Director of Product Marketing, as well as Neeraj Varma, Country Manager, Sales, for India and Australia and New Zealand.
First up, the ISE Design Suite 11.2 is now available for download, with full public support for Virtex-6 and Spartan-6 FPGA families. Xilinx also introduced the Spartan-6 and Virtex-6 base evaluation kits, which can be order immediately by customers.
Since this is the ISE Design Suite 11.2, Przybus added that prior to June 24, support for the Virtex-6 and Spartan-6 FPGA families was available only to early access customers.
Accelerating development of SoC solutions
Next, the Base Targeted Design Platform is said to accelerate the development of SoC solutions. According to Przybus, the Base Targeted Design Platform provides a framework that customers can extend to build their SoC solutions.
"We are providing the common functions including host interface and external memory controller as well as multi-boot in a system configuration. Customers can leverage this code saving weeks of months of development."
Why now?
The obvious question, why this release now, and not earlier? According to Varma: "Xilinx has always had boards, silicon, tools, IP and reference designs. However, with changes in market conditions and customer needs evolving, what became abundantly clear in recent years is that we need a more formalized and efficient way of providing a base for customers to build upon. Customers have also been asking for more complete design solutions."
The concept for Base Targeted Design Platform was introduced in February when Xilinx had announced the architectural details of Spartan-6 and Virtex-6 FPGA devices along with the entire targeted design platform strategy.
"When the announcement was made, we had early access customers designing with Spartan-6 and Virtex-6 FPGAs. With the release of the ISE Design Suite 11.2, we are opening up public access to the two new device families. By doing so, we are opening up access through software of all our new devices, we are also making technical documentation, user guides and other resources available to all customers," he added.
The release of the Base Targeted Design Platform is coincidental to the public availability of software supporting both Virtex-6 and Spartan-6 FPGAs.
The new Virtex-6 FPGA and Spartan-6 FPGA Evaluation Kits are the first in a series of kits that Xilinx will offer throughout the year designed to simplify the evaluation and development of SoCs with the latest generation of programmable technologies from Xilinx.
Now that the first kits have been released, let us probe into Xilinx's plan for evaluation kits that it will offer throughout the year.
Varma added: "According to our Targeted Design Platform strategy, we have introduced the first level of our offerings. Moving forward, throughout the year we will introduce the Domain Specific Platform and then the Market Specific Platform. The Domain Kits will incorporate embedded kits, connectivity kits, and finally the DSP kits for both Virtex-6 and Spartan-6." This point should be noted with great care by designers as lots more is in the offing from Xilinx.
The Market Specific Platform will address specific markets and include Communication, Video and Broadcast Kits, Market specific IP, custom tools and custom boards, added Varma.
Spartan-6 SP601 evaluation kit
Xilinx also introduced the Spartan-6 SP601 evaluation kit. Brent Przybus highlighted that the Spartan-6 SP601 evaluation kit is designed to address customers developing high volume, lower cost applications.
He elaborated: "The kit features the Spartan-6 LX16 FPGA and ships with a full base reference design and interface software providing customers a host communications link, built-in memory controller core that interfaces to DDR2 DRAM on the SP601 board, support for multi-boot, and a processing block that enables customers to see and measure the benefits of using a hard IP vs. Logic only simple processing function.
Addressing defense, aerospace apps
How useful will be all of this for defense and aerospace applications? According to Neeraj Varma, a lot of aerospace and defense applications require high performance digital signal processing (DSP), for example, in their video processing, secure communications, wireless communications (software defined radio or SDR), etc.
"Using the Base Targeted Design PLtatform as shown in the demonstration video, designers will be able to evaluate tradeoffs in performance, precision, and power consumption using hard DSP slices available in Spartan-6 and Virtex-6 FPGAs," he said.
Using the DSP slices in Spartan-6 will help designers boost their performance by five times with higher precision without resulting into the increase of overall power consumption.
Varma added: "The Base Targeted Platform that we have announced will help our customers tune their applications not only in the aerospace and defence applications, but can be used for other applications in different vertical markets as well. Through this year, we will introduce Domain Specific Kits followed by Market Specific Kits. The Market Specific Platform will further address the specific defense applications in future."
Lastly, there has been a lot of focus on design re-engineering and design security.
Przybus pointed out that the specific reference design shown in the Xilinx demo video has been done in HDL and doesn't include and design security.
"The base reference design is portable and can be extended in a number of ways. The customer could use the built-in features like DeviceDNA, AES encryption of bitstream, etc. in our Virtex-6 and Spartan-6 silicon to secure their designs," he noted.
I was in conversation with Brent Przybus, Director of Product Marketing, as well as Neeraj Varma, Country Manager, Sales, for India and Australia and New Zealand.
First up, the ISE Design Suite 11.2 is now available for download, with full public support for Virtex-6 and Spartan-6 FPGA families. Xilinx also introduced the Spartan-6 and Virtex-6 base evaluation kits, which can be order immediately by customers.
Since this is the ISE Design Suite 11.2, Przybus added that prior to June 24, support for the Virtex-6 and Spartan-6 FPGA families was available only to early access customers.
Accelerating development of SoC solutions
Next, the Base Targeted Design Platform is said to accelerate the development of SoC solutions. According to Przybus, the Base Targeted Design Platform provides a framework that customers can extend to build their SoC solutions.
"We are providing the common functions including host interface and external memory controller as well as multi-boot in a system configuration. Customers can leverage this code saving weeks of months of development."
Why now?
The obvious question, why this release now, and not earlier? According to Varma: "Xilinx has always had boards, silicon, tools, IP and reference designs. However, with changes in market conditions and customer needs evolving, what became abundantly clear in recent years is that we need a more formalized and efficient way of providing a base for customers to build upon. Customers have also been asking for more complete design solutions."
The concept for Base Targeted Design Platform was introduced in February when Xilinx had announced the architectural details of Spartan-6 and Virtex-6 FPGA devices along with the entire targeted design platform strategy.
"When the announcement was made, we had early access customers designing with Spartan-6 and Virtex-6 FPGAs. With the release of the ISE Design Suite 11.2, we are opening up public access to the two new device families. By doing so, we are opening up access through software of all our new devices, we are also making technical documentation, user guides and other resources available to all customers," he added.
The release of the Base Targeted Design Platform is coincidental to the public availability of software supporting both Virtex-6 and Spartan-6 FPGAs.
The new Virtex-6 FPGA and Spartan-6 FPGA Evaluation Kits are the first in a series of kits that Xilinx will offer throughout the year designed to simplify the evaluation and development of SoCs with the latest generation of programmable technologies from Xilinx.
Now that the first kits have been released, let us probe into Xilinx's plan for evaluation kits that it will offer throughout the year.
Varma added: "According to our Targeted Design Platform strategy, we have introduced the first level of our offerings. Moving forward, throughout the year we will introduce the Domain Specific Platform and then the Market Specific Platform. The Domain Kits will incorporate embedded kits, connectivity kits, and finally the DSP kits for both Virtex-6 and Spartan-6." This point should be noted with great care by designers as lots more is in the offing from Xilinx.
The Market Specific Platform will address specific markets and include Communication, Video and Broadcast Kits, Market specific IP, custom tools and custom boards, added Varma.
Spartan-6 SP601 evaluation kit
Xilinx also introduced the Spartan-6 SP601 evaluation kit. Brent Przybus highlighted that the Spartan-6 SP601 evaluation kit is designed to address customers developing high volume, lower cost applications.
He elaborated: "The kit features the Spartan-6 LX16 FPGA and ships with a full base reference design and interface software providing customers a host communications link, built-in memory controller core that interfaces to DDR2 DRAM on the SP601 board, support for multi-boot, and a processing block that enables customers to see and measure the benefits of using a hard IP vs. Logic only simple processing function.
Addressing defense, aerospace apps
How useful will be all of this for defense and aerospace applications? According to Neeraj Varma, a lot of aerospace and defense applications require high performance digital signal processing (DSP), for example, in their video processing, secure communications, wireless communications (software defined radio or SDR), etc.
"Using the Base Targeted Design PLtatform as shown in the demonstration video, designers will be able to evaluate tradeoffs in performance, precision, and power consumption using hard DSP slices available in Spartan-6 and Virtex-6 FPGAs," he said.
Using the DSP slices in Spartan-6 will help designers boost their performance by five times with higher precision without resulting into the increase of overall power consumption.
Varma added: "The Base Targeted Platform that we have announced will help our customers tune their applications not only in the aerospace and defence applications, but can be used for other applications in different vertical markets as well. Through this year, we will introduce Domain Specific Kits followed by Market Specific Kits. The Market Specific Platform will further address the specific defense applications in future."
Lastly, there has been a lot of focus on design re-engineering and design security.
Przybus pointed out that the specific reference design shown in the Xilinx demo video has been done in HDL and doesn't include and design security.
"The base reference design is portable and can be extended in a number of ways. The customer could use the built-in features like DeviceDNA, AES encryption of bitstream, etc. in our Virtex-6 and Spartan-6 silicon to secure their designs," he noted.
Monday, July 6, 2009
Union budget 2009: Nothing much to speak about on tech front, barring UIDAI!
The Union Budget 2009, detailed today by Honourable Pranab Mukherjee, Minister of Finance, Government of India, really has nothing much to speak about for the Indian technology sector, or what many would like to call as the Indian IT industry, barring the setting up of the Unique Identification Authority of India (UIDAI).
Some of the budget highlights include:
* Customs duty of 5 percent to be imposed on set-top boxes for TV broadcasting.
* Customs duty on LCD panels for manufacture of LCD TVs to be reduced from 10 percent to 5 percent.
* Full exemption from 4 percent special CVD on parts for manufacture of mobile phones and accessories to be reintroduced for one year.
None of these proposals will significantly boost manufacturing in the country. There is also a great need to boost home-grown companies!
Among the good points, the minister said that he will urge his colleagues in Central and State Governments to remove policy, regulatory and institutional bottlenecks for speedy implementation of infrastructure projects. The infrastructure projects include telecommunications, power generation, etc. This is some good news!
The IT industry has pointed out that it is facing difficulties in the assessment of software which involves transfer of the right to use after the levy of service tax on IT software service. To resolve the matter, the minister has proposed to exempt the value attributable to the transfer of the right to use packaged software from excise duty and CVD. Perhaps, this is some good news as well.
The minister also noted that the setting up of the Unique Identification Authority of India (UIDAI) is a major step in improving governance with regard to delivery of public services. The first set of unique identity numbers will be rolled out in 12 to 18 months. Rs. 120 crore has been provided for this project. This is also good news.
One other point to note is the feduction in respect of export profits available under sections 10A and 10B of the Income-tax Act. The deduction under these sections would not be available beyond the financial year 2009-2010. To tide over the slowdown in exports, the minister has proposed to extend the sun-set clauses for these tax holidays by one more year, i.e., for the financial year 2010-11.
Semicon, solar seem neglected this time!
The budget has missed out in helping develop the semiconductor and solar/PV industries. These sectors require the full backing of the government. Even local telecom manufacturing seems to have been bypassed. Nor is there any mention of how foreign direct investment (FDI), can be enhanced in these critical sectors.
Especially, solar photovoltaics holds a lot of promise. A couple of months ago, SEMI India, in its paper on solar PV in India, had highlighted the need for more action from the government of India, a more closer industry-government collaboration, as well as the need for financial institutions to pay more attention to the solar/PV segment in India.
I also didn't see any proposal in the budget that would help strengthen India's semiconductor ecosystem. How can India become an even more attractive destination for foreign investors? There are companies, especially some Indian technocrats, who would probably like to return to India and set up semiconductor product companies. Would the VC community finance semicon start-ups? What would excite these folks?
Recently, BV Naidu, chairman, India Semiconductor Association, (ISA), mentioned in a discussion, that it would be appropriate if the Government of India could provide seed and start-up capital for new ventures and set up a focused venture fund of about Rs 200 crores.
According to him, the technology development board could administer these funds and the fund may provide up to 80 percent of the approved project cost with equity balance being brought in by entrepreneur. The Government can also subsidize the acquisition of EDA tools by start-ups and other SMEs in this sector.
None of these points have been addressed in today's union budget!
Someone recently questioned me whether India will ever have a fab. Frankly, I don't know! Having a good, strong local semiconductor industry does require a huge support from the government.
This could've been a much more braver and bolder budget. Perhaps, future budgets will address points given a miss in this edition.
Some of the budget highlights include:
* Customs duty of 5 percent to be imposed on set-top boxes for TV broadcasting.
* Customs duty on LCD panels for manufacture of LCD TVs to be reduced from 10 percent to 5 percent.
* Full exemption from 4 percent special CVD on parts for manufacture of mobile phones and accessories to be reintroduced for one year.
None of these proposals will significantly boost manufacturing in the country. There is also a great need to boost home-grown companies!
Among the good points, the minister said that he will urge his colleagues in Central and State Governments to remove policy, regulatory and institutional bottlenecks for speedy implementation of infrastructure projects. The infrastructure projects include telecommunications, power generation, etc. This is some good news!
The IT industry has pointed out that it is facing difficulties in the assessment of software which involves transfer of the right to use after the levy of service tax on IT software service. To resolve the matter, the minister has proposed to exempt the value attributable to the transfer of the right to use packaged software from excise duty and CVD. Perhaps, this is some good news as well.
The minister also noted that the setting up of the Unique Identification Authority of India (UIDAI) is a major step in improving governance with regard to delivery of public services. The first set of unique identity numbers will be rolled out in 12 to 18 months. Rs. 120 crore has been provided for this project. This is also good news.
One other point to note is the feduction in respect of export profits available under sections 10A and 10B of the Income-tax Act. The deduction under these sections would not be available beyond the financial year 2009-2010. To tide over the slowdown in exports, the minister has proposed to extend the sun-set clauses for these tax holidays by one more year, i.e., for the financial year 2010-11.
Semicon, solar seem neglected this time!
The budget has missed out in helping develop the semiconductor and solar/PV industries. These sectors require the full backing of the government. Even local telecom manufacturing seems to have been bypassed. Nor is there any mention of how foreign direct investment (FDI), can be enhanced in these critical sectors.
Especially, solar photovoltaics holds a lot of promise. A couple of months ago, SEMI India, in its paper on solar PV in India, had highlighted the need for more action from the government of India, a more closer industry-government collaboration, as well as the need for financial institutions to pay more attention to the solar/PV segment in India.
I also didn't see any proposal in the budget that would help strengthen India's semiconductor ecosystem. How can India become an even more attractive destination for foreign investors? There are companies, especially some Indian technocrats, who would probably like to return to India and set up semiconductor product companies. Would the VC community finance semicon start-ups? What would excite these folks?
Recently, BV Naidu, chairman, India Semiconductor Association, (ISA), mentioned in a discussion, that it would be appropriate if the Government of India could provide seed and start-up capital for new ventures and set up a focused venture fund of about Rs 200 crores.
According to him, the technology development board could administer these funds and the fund may provide up to 80 percent of the approved project cost with equity balance being brought in by entrepreneur. The Government can also subsidize the acquisition of EDA tools by start-ups and other SMEs in this sector.
None of these points have been addressed in today's union budget!
Someone recently questioned me whether India will ever have a fab. Frankly, I don't know! Having a good, strong local semiconductor industry does require a huge support from the government.
This could've been a much more braver and bolder budget. Perhaps, future budgets will address points given a miss in this edition.
Subscribe to:
Posts (Atom)