Early
this month, I hobbled along to catch up with Jaswnder Ahuja, corporate
VP and MD, Cadence Desiign Systems India. With the global semiconductor
industry having entered the sub-20nm era, there are a lot of things
happening, and Cadence is sure to be present.
First, let's see how's the global semiconductor industry performing after entering the sub-20nm era.
Ahuja
replied: "Increased demand for faster, smaller, low-power chips
continues to drive the geometry shrink as one of the ways to manage the
low-power, higher performance goals in smaller form factors—in other
words, PPA is driving the move to advanced node design.
"At
Cadence, we are seeing a lot of interest in the wireless space, which
includes smartphones, tablets, and consumer devices. In this market, you
must support different standards, the device must be really fast, it
must have Internet access, and all this must be done at lower power so
the that it does not drain the battery. We’re also seeing interest for
advanced nodes in other segments such as computing and graphics
processors."
When speaking of advanced nodes, let's also try and
find out what Cadence is doing in helping achieve 10X faster power
integrity analysis and signoff.
Cadence Voltus IC power
integrity Solution is a full-chip, cell-level power signoff tool that
provides accurate, fast, and high-capacity analysis and optimization
technologies to designers for debugging, verifying, and fixing IC chip
power consumption, IR drop, and electromigration (EM) constraints and
violations.
The Voltus solution includes innovative technologies
such as massively parallel execution, hierarchical architecture, and
physically aware power grid analysis and optimization. Beneficial as a
standalone power signoff tool, Voltus IC Power Integrity Solution
delivers even more significant productivity gains when used in a highly
integrated flow with other key Cadence products, providing the
industry’s fastest design closure technology.
Developed with
advanced algorithms and a new power integrity analysis engine with
massively parallel execution, Voltus IC Power Integrity Solution:
* Performs 10X faster than other solutions on the market.
* Supports very large designs—up to one billion instances—with its hierarchical architecture.
* Delivers SPICE-level accuracy.
* Enhances physical implementation quality via physically aware power integrity optimization.
Supported
by major foundries and intellectual property (IP) providers, Voltus IC
Power Integrity Solution has been validated and certified on advanced
nodes processes such as 16nm FinFET and included in reference design
flows such as for 3D-IC technology.
Backed by Cadence’s rigorous
quality control and product release procedures, the Voltus solution
delivers best-in-class signoff quality on accuracy and stability for all
process nodes and design technologies.
FinFETs to 20nm - are folks benefiting?
It is common news that
FinFETs have gone to 20nm and perhaps, lower. Therefore, are those folks
looking for power reduction now benefiting?
Ahuja replied that
FinFETs allow semiconductor and systems companies to continue to develop
commercially viable chips for the mobile devices that are dominating
the consumer market. FinFETs enable new generations of high-density,
high-performance, and ultra-low-power systems on chip (SoCs) for future
smart phones, tablets, and other advanced mobile devices. Anyone who
adopts FinFET technology will reap the benefits.
Foundry support
for FinFETs will begin at 16nm and 14nm. In April of this year, Cadence
announced a collaboration with ARM to implement the industry’s first ARM
Cortex-A57 processor on TSMC’s 16nm FinFET manufacturing process. At
ARM TechCon 2012, Cadence announced a 14nm test chip tapeout using an
ARM Cortex-M0 processor and IBM's FinFET process technology.
Meeting design challenges at 20nm
We
will also check out how's Cadence meeting the design challenges at
20nm, with respect to the need for double patterning, to overcome the
limits of existing lithography equipment.
Ahuja noted that there
are several challenges associated with double patterning. Double
patterning impacts cell and library generation. You must make sure
silicon intellectual property (IP) is compliant with double patterning
layout rules. It is also critical to account for double patterning
during placement.
Cadence’s unique technology performs automatic
colorized placement, with the end benefit being a less congested
design. With less congestion, it is much easier to meet timing and power
requirements.
The biggest impact of double patterning is in
routing. The double patterning must be integrated inside the routing
solution—it cannot be an afterthought where you finish the routing and
then run decomposition. It must be correct-by-construction.
Cadence’s
approach is to carry double patterning intent forward, from cell and IP
generation to double-pattern aware routing, and finally, to signoff
physical verification. This approach provides faster convergence because
intent is carried forward throughout the flow. A second benefit is
better quality of results.
Challenges like double patterning,
clock design, and layout-dependent effects all must be considered up
front in the design flow, from IP characterization to placement and
routing and final signoff. To confront these challenges, an end-to-end
flow is required, and that is what Cadence offers.
So, what is
Cadence doing in 20nm that involves in-design physical verification
throughout the flow to reduce time-consuming, uncertain iterations?
In-design
physical verification is becoming more complex at advanced nodes when
designs are bigger and the number and complexity of DRC rules are more.
Ensuring a DRC clean design while still meeting the tapeout window is a
challenge.
Proven on many successful production tapeouts in
nanometer process technologies, Cadence Physical Verification System is
the premier Cadence signoff solution enabling in-design and back-end
physical verification, constraint validation, and reliability checking.
Recently
PMC adopted the Cadence Physical Verification System as the signoff
technology for its global design centers. PMC has used the Physical
Verification System for several successful tapeouts, including PMC’s
DIGI 120, described as the industry’s only single-chip processor
supporting 10G, 40G, and 100G speeds for OTN transport, aggregation, and
switching. The device, with 200+ million gates and 180+ Mbits of RAM,
is the largest production SoC that PMC has delivered.
Next, it would be interesting to see how Cadence has been handling the early RTL and design explorations.
Ahuja
said that accommodating today’s chip design requirements within narrow
market windows has led to a predictability crisis. How can engineers
determine design feasibility for larger, higher performance,
power-hungry chips with an incomplete netlist, library, and constraints?
And, how can they quickly assess floorplans for congestion, timing, and
power without having to go into real implementation?
Cadence
provides two solutions. The first is Cadence First Encounter Design
Exploration and Prototyping technology. First Encounter offers a
comprehensive flat and hierarchical design planning, analysis, and debug
environment for complex designs. Its GigaFlex technology adapts to
growing capacity requirements while still retaining the relevant timing,
placement, and congestion information to accurately plan and implement
100M+ instance designs.
The latest advanced node and low-power
capabilities make hierarchical implementation of high-speed and
advanced-node designs faster and easier, which in turn produces designs
that consume lower power. In addition, with unique partitioning and
budgeting capabilities, First Encounter technology ensures convergence
and provides a predictable path to design closure.
The second
solution is Cadence InCyte Chip Estimator. Incyte enables rapid and
accurate early chip planning. Fast “what-if” estimation of size,
performance, power, and cost enables early exploration of functional
content, IP components, memories, and process technology with foresight
into the technical and business consequences of these decisions. InCyte
Chip Estimator helps chip design projects start and stay on track toward
their technical and business goals.
Status of 3D ICs
Finally, what's the status with 3D ICs? How is Cadence helping with true 3D stacking integration?
Performance
and area real estate have been driving the vertical dimension for
several years. 3D IC packages pack heterogeneous die—logic, memory, RF,
MEMs—at different process nodes to create different solutions. Micron is
bullish on its Hybrid Memory Cube technology. SK Hynix, Fujitsu,
Samsung, and others are said to have technologies in the works. TSMC,
GlobalFoundries, and others have 2.5D offerings for customers.
From
a design standpoint, extensive retooling is not needed for 3D ICs.
Cadence offers a comprehensive solution to support the 3D IC revolution,
including analog and digital implementation, packaging, and printed
circuit board (PCB) design tools. And there are no apparent showstoppers
in process technology.
Ahuja noted that in September 2013,
Cadence announced its partnership with TSMC to develop a 3D-IC reference
flow featuring innovative true 3D stacking. The flow, validated on a
memory-on-logic design with a 3D stack based on a Wide I/O interface,
enables multiple die integration, incorporating TSMC 3D stacking
technology and Cadence® solutions for 3D-IC, including integrated
planning tools, a flexible implementation platform, and signoff and
electrical/thermal analysis.
Sunday, December 22, 2013
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.