Archive

Archive for the ‘28nm’ Category

FinFETs delivering on promise of power reduction: Synopsys

February 1, 2014 Comments off

Here is the concluding part of my conversation with Synopsys’ Rich Goldman on the global semiconductor industry.

Rich Goldman

Rich Goldman

Global semicon in sub 20nm era
How is the global semicon industry performing after entering the sub 20nm era? Rich Goldman, VP, corporate marketing and strategic alliances, Synopsys, said that driving the fastest pace of change in the history of mankind is not for the faint of heart. Keeping up with Moore’s Law has always required significant investment and ingenuity.

“The sub-20nm era brings additional challenges in device structures (namely FinFETs), materials and methodologies. As costs rise, a dwindling number of semiconductor companies can afford to build fabs at the leading edge. Those thriving include foundries, which spread capital expenses over the revenue from many customers, and fabless companies, which leverage foundries’ capital investment rather than risking their own. Thriving, leading-edge IDMs are now the exception.

“Semiconductor companies focused on mobile and the Internet of Things are also thriving as their market quickly expands. Semiconductor companies who dominate their space in such segments as automotive, mil/aero and medical are also doing quite well, while non-leaders find rough waters.”

Performance of FinFETs
Have FinFETs gone to below 20nm? Also, are those looking for power reduction now benefiting?

He added that 20nm was a pivotal point in advanced process development. The 20nm process node’s new set of challenges, including double patterning and very leaky transistors due to short channel effects, negated the benefits of transistor scaling.

To further complicate matters, the migration from 28nm to 20nm lacked the performance and area gains seen with prior generations, making it economically questionable. While planar FET may be nearing the end of its scalable lifespan at 20nm, FinFETs provide a viable alternative for advanced processes at emerging nodes.

The industry’s experience with 20nm paved the way for an easier FinFET transition. FinFET processes are in production today, and many IC design companies are rapidly moving to manufacture their devices on the emerging 16nm and 14nm FinFET-based process geometries due to the compelling power and performance benefits. Numerous test chips have taped out, and results are coming in.

“FinFET is delivering on its promise of power reduction. With 20nm planar FET technologies, leakage current can flow across the channel between the source and the drain, making it very difficult to completely turn the transistor off. FinFETs provide better channel control, allowing very little current to leak when the device is in the “off” state. This enables the use of lower threshold voltages, resulting in better power and performance. FinFET devices also operate at a lower nominal voltage supply, significantly improving dynamic power.”
Read more…

Dr. Wally Rhines: Watch out for 14/16nm technologies in 2014!

December 6, 2013 Comments off

Dr. Wally RhinesIt is always a pleasure speaking with Dr. Walden (Wally) C. Rhines, chairman and CEO, Mentor Graphics Corp. The last time I met him was at Santa Cruz, USA, during a global electronics forum in April this year. First, I asked him regarding the outlook for the global semiconductor industry in 2014, as well as the EDA industry.

Outlook for global semicon industry in 2014
Dr. Rhines said: “The outlook for the global semicon industry in 2014 is modestly positive. Most analysts will see single digiit growth. In memory, we have short supply vs. demand. While we had consolidation of the wireless industry, we still have volumes of handsets, tablets, etc. In the US, tablets are said to be the biggest growth area during Xmas.

“When you look at any product, you look at what more can it do. You look at more and more features that can be added. We have speciallization in ARM-based chips. There are enough change dynamics that show demand. The iPad bridged the gap between the portable PC and phone. The infrastructure of apps has now made a huge infrastructure. If you are dependant on apps, there can be a differentiator.

“Wearable electronics is another great opportunity. However, it is still a small market. The electronic watch is interesting. We are in an era where there are some things that are key, and some require figuring out. There will be more and more need for specific devices, rather than only applications in future. The same thing was with the PC, which went from custom to specific needs.”

In that case, how is the global semiconductor industry performing having entering the sub 20nm era?

He said that 2014 is going to be a big year. There will be releases of 14/16nm technologies. This will be the year when customers will be doing tests. There are companies in all regions of the world that will be doing such stuff.

Have FinFETs gone to 20nm? Are those looking for power reduction now benefiting?

Dr. Rhines said: “The big advantage is leakage. FinFET dramatically impacts current leakage. Now, attention will shift to dynamic power. It will once again be predominantly the consumer of power in large chips.

Outlook for EDA industry
Now, let’s see what’s the outlook for the global EDA industry in 2014.

Dr. Rhines said: “Whenever you create new technologies, you will need EDA. So, EDA will grow. New designs will also need EDA. There will be new EDA tools. EDA is now addressing thermal and stress issues in verification and design. Caliber PERC is our main product here. The upgrades are good for EDA. There are new things they have to adopt, in these tools.”

Let’s talk a bit about embedded. Mentor released the new version of Sourcery CodeBench. What does it stand to gain?

Raghu Panicker, sales director, Mentor Graphics India said the Sourcery CodeBench is a real-time operating system (RTOS). That product is gaining momentum. Large MNC customers like Qualcomm are adopting this. Among small firms, there are medical, energy meter companies that are handling it as well.

Dr. Rhines added that Sorcery CodeBench is indicative of a trend – it is very open source based. It is now 20,000 downloads a month, so that is a big community.

Next is there any scope for the growth of biomems and optical telecom industry?

He said that both areas are interesting. Biomems are still a fairly small market. It is going to be evolutionary. As for optical telecom, over the last year or two, all participants have gone into a silent mode. Mentor is working with a number of customers.

Five trends to rule in 2014
Now, it was quiz time. First, the top five trends in the EDA industry during 2014. Dr. Rhines said:
* Growth of emulation for verification. The market is growing at over CAGR of 25 percent. Emulation is really big. It will be a big game changer for EDA.
* 16/14nm.
* Continued pressure on power as we go to FinFETs.
* Power reduction.
* Yield analysis for 14/16nm. A near range can be security.

Now, the top five trends for semiconductors in 2014! Dr. Rhines mentioned these as:
* Move to 14/16nm and cost.
* Growth in hybrid functions is another trend.
* Basic IoT.
* Security – how you verify designs.
* Continued commoditization of wireless apps.
Read more…

Outlook 2014: Xilinx bets big on 28nm

December 4, 2013 Comments off

Neeraj Varma

Neeraj Varma

We are in December, and its time for outlook 2014! First, I met up with Neeraj Varma, director-Sales, India, Xilinx. He said: “We expect the 28nm to do really well. From Apr. 13-Mar. 14, we expect revenues worth $250 milion from the 28nm line.

“We are now looking at the embedded market – and expect about $2 billion serviceable available market (SAM). We are looking at $8 billion SAM at the ASIC/ASSP displacement market, and of course $6 billion SAM for core PLD.” After a long time, Xilinx has been seeing positive capex. “We are entering a growth cycle for service providers and enterprises,” he added.

A macro view of capex equipment spend is driven by LTE 27.2 percent at 2011-16, and optical networks 15.9 percent. The other areas include data center, enterprise switching and routing, and service provider switching and routing. Next, 3D ICs will enable Nx100G OTN, 400G OTN, MuxSAR, as well as top of the rack switch, I/O virtualization.

Earlier, there were less than 50 ASICs start in communications in the top 10 OEMs. There were less than 20 28nm ASIC starts in at top 10 OEMs. As of 2012, less than 50 percent of the top 16 ASSPs vendors were losing money. Customer needs are diverse now. Companies end up over designing a chip. People end up paying for what trey are not using.

Xilinx is offering the SMARTCORE IP for smarter networks and data centers. “40 percent of our wins have been achieved by integrating or displacing ASICs and ASSPs,” he said. “We have 25 percent total wins across a broad set of apps/portfolio.”

Some other gains for Xilinx:
* Xilinx gained 3 percent increase in PLDs.
* In wired and data centers, it has 12-percent CAGR from 2013-16.
* In wireless, it has 10-12 percent CAGR.
* In automotive smarter vision, it has 20 percent CAGR growth.
* In industrial, scientific and medical (ISM), it has 12 percent CAGR growth.
* In FY13E-FY16E, Xilinx expects to grow 8-12 percent, and has plans to increase the R&D revenue to 8.6 percent.

Great, India’s having fabs! But, is the tech choice right?

September 13, 2013 2 comments

G450C

G450C

The government of India recently approved the setting up of two semiconductor wafer fabrication facilities in the country. It is expected to provide a major boost to the Indian electronics system design and manufacturing (ESDM) ecosystem. A look at the two proposals:

Jaiprakash Associates, along with IBM (USA) and Tower Jazz (Israel). The outlay of the proposed fab is about Rs. 26,300 crore for establishing the fab facility of 40,000 wafer starts per month of 300mm size, using advanced CMOS technology. Technology nodes proposed are 90nm, 65nm and 45nm nodes in phase I, 28nm node in phase II with the option of establishing a 22nm node in phase III. The proposed location is Greater Noida.

Hindustan Semiconductor Manufacturing Corp. (HSMC) along with ST Microelectronics (France/Italy) and Silterra (Malaysia). The outlay of the proposed fab is about Rs. 25,250 crore for the fab facility of 40,000 wafer starts per month of 300mm size, using advanced CMOS technology. Technology nodes proposed are 90nm, 65nm and 45nm nodes in phase I and 45nm, 28nm and 22nm nodes in phase II. The proposed location is Prantij, near Gandhinagar, Gujarat.

Now, this is excellent news for everyone interested in the Indian semiconductor industry.

One look at the numbers above tell me – NONE OF THESE are going to be 450mm fabs! Indeed, both will be 300mm fabs! After waiting for such a long time to even get passed by the Union Cabinet, are these 300mm fabs going to be enough for India? Is the technology choice even right for the upcoming wafer fabs in India? Let’s examine!

As you can probably see, both the projects have placed 22nm right at the very last phase! That’s very interesting!

Intel just showcased its Xeon processor E5-2600 v2 product family a few days back. I distinctly remember Intel’s Narendra Bhandari showing off the 22nm wafer sometime last week during a product launch!

For discussion’s sake, let’s say, a fab in India comes up by say, early 2015. Let’s assume that Phase 1 takes a full year. Which means, Phase 2, where 22nm node would be used, shall only be touched in 2016 or even beyond! Isn’t it? Where will the rest of the global industry be by then?

You are probably aware of the Global 450 Consortium or G450C, which has Intel, IBM, Samsung, GlobalFoundries and TSMC among its members.  What is the consortium currently doing? It is a 450mm wafer and equipment development program, which is leveraging on the industry and government investments to demonstrate 450mm process capabilities at the CNSE’s Albany Nanotech Complex. CNSE, also a consortium member, is the SUNY’s College of Nanoscale Science and Engineering!

So, what does all of this tell me?

One, these upcoming fabs in India will probably produce low- to mid-range chips, and some high-end ones at a later stage. Well, two, this does raise a question or two about India’s competitive advantage in the wafer fab space!  Three, there is lot of material on 450mm fabs, and some of that is available right here, on this blog! Have the Indian semiconductor industry folks paid enough attention to all that? I really have no idea!

Four, only the newer 300mm fabs built with higher ceilings and stronger floors will be able to be upgraded to 450mm, as presented by The Information Network’s Dr. Robert Castellano at the Semicon West 2013. Five, what are the likely alternative markets for 200mm and 300mm fabs? These are said to be MEMs and TSV, LEDs and solar PV. Alright, stop!

Perhaps, these product lines will be good for India and serve well, for now, but not for long!

Xilinx tapes-out first UltraScale ASIC-class programmable architecture

July 19, 2013 Comments off

Xilinx.

Xilinx’s UltraScale architecture.

Xilinx Inc. has taped-out the first 20nm All Programmable Device with first UltraScale ASIC-class programmable architecture. It is said to be the semiconductor industry’s first 20nm device, and the PLD industry’s first 20nm All Programmable device. Xilinx implemented the industry’s first ASIC-class programmable architecture called UltraScale.

These milestones expand on Xilinx’s industry first 28nm tape-out, All Programmable SoCs, All Programmable 3D ICs, and SoC-strength design suite. Xilinx already has several firsts in the 28nm space, such as:

* First 28nm tape-out.
* First All Programmable SoC.
* First All Programmable 3D IC.
* First SoC-strength design suite.

Neeraj Varma, director-Sales, India, said that Xilinx’s global market share in the 28nm portfolio was 65 percent in March 2013. With the launch of the industry’s first 20nm All Programmable Device with first UltraScale ASIC-class programmable architecture, there are improvements such as 1.5-2x performance and integration, and a year ahead of the competition. It handles massive I/O bandwidth, massive memory bandwidth, massive data flow and routing, and fastest DSP processing. The architecture will scale — from monolithic to 3D IC, planar to FinFET, and ASIC-class performance.

The UltraSCALE architecture points to high performance smarter systems. For example, 1Tps in OTN networking, 8K in digital video, LTE-A in wireless communications, and digital array in radar. There will be requirements for massive packet processing over 400 Gbps wire-speed, massive data flow over 5Tbps, as well as massive I/O and memory bandwidth over 5Tbps, and DSP performance over 7 TMACs.

The mandate for ASIC-class programmable architecture is to remove bottlenecks for massive data flow and smart processing, high throughput with low latency, and efficient design closure with greater than 90 percent utilization without performance degradation. These are the benefits of applying leading edge ASIC techniques in a fully programmable architecture.

ASIC-like clocking maximizes performance margin for highest throughput. UltraSCALE ASIC-like clocking enables clock placement virtually anywhere on the die, making the clock skew problem go away. Also, highly optimized critical paths remove bottlenecks in DSP and packet processing. There is greatly enhanced DSP processing, high-speed memory cascading, and hardened IP for I/O intensive functions.

Next generation power management features also enable a leap in performance. The process node is up to 35 percent static at 20nm. There are more buffers for granular or coarse clock gating. Block RAM is dynamic power gating, hardened cascading. For transceivers, there are architectural optimizations. There is efficient packing and utilization of the logic fabric. For DSP, there are wider multipliers and fewer blocks per function. As for memory, there is DDR4, which operates at 1.2v vs.1.5v, voltage scaling.

The Xilinx KINTEX UltraSCALE will power 4×4 mixed-mode radios, 100G traffic manager NICs, super high-vision processing, 256-channel ultrasound and 48-channel T/R radar processing. The Xilinx VIRTEX UltraSCALE will power 400G OTN switching, 400G transponder, 400G MAC-to-Interlaken bridge, 2x100G muxponder and ASIC prototyping.

Xilinx worked with TSMC to infuse high-end FPGA requirements into the TSMC 20SoC development process, just as it had done in the development of 28HPL. The Xilinx Vivado Design Suite early access supporting UltraScale architecture-based FPGAs is now available. Initial UltraScale devices will be available in Q4-2013.

Xilinx announces 20nm portfolio strategy

November 14, 2012 Comments off

Source: Xilinx, USA.

Source: Xilinx, USA.

Xilinx Inc. has announced its 20nm portfolio strategy. The 20nm portfolio will allow Xilinx to offer twice the performance at half the power. It will increase productivity by 4x, and improve integration by 1.5- 2x. Besides, there will be 20-50 percent lower BOM cost.

Xilinx’s 20nm all programmable portfolio builds on 28nm breakthroughs to stay a generation ahead.  “At 20nm, we were able to break out to become an all programmable company,” said Neeraj Varma, country manager, sales, India, Australia and New Zealand, Xilinx India.

The next generation FPGAs, second generation SoCs and 3D ICs will be ‘co-optimized’ with Vivado for the most compelling alternative ever to ASICs and ASSPs. From enabling programmable logic, the Xilinx 20nm portfollio will enable programmable systems integration!

The first SoC strength design suite was shipped in Q2-2012. It has been built from ground up for the next decade of all programmable devices. Today, the Xilnix Vivado is used for over 30 percent of 28nm FPGAs and 100 percent for 3D ICs.

Xilinx has been expanding on its next generation competencies. The 3D IC expertise and supply chain has gone from homogenous to heterogenous. The SoC and embedded software has also undergone change, as have XCVRs and analog mixed signal (AMS), communications BU and applications IP, and next generation design automation. Xilinx is now charting an aggressive course forward.

Xilinx’s 20nm portfolio has been co-optimized for performance, power and integration to address the market needs at 20nm. For the next-generation FPGA,, it will provide unmatched system optimized transceivers at highest channel quality w/ second generation auto equalization. There will be higher bandwidth w/over 100 transceivers @ 33Gb/s.

There will be 2X performance optimization, with faster DSP and BRAM, DDR4, transceivers and 2x memory bandwidth. There will be over 90 percent routing architecture enabling high bandwidth bussing and fast design. One half power optimization will provide an optimized performance/watt. There will be next generation block level power management. There will be 1.5x integration/BOM in terms of 1.5x logic, DSP, BRAM, AMS, VCXO, etc. Read more…

Realizing EDA360: Charlie Huang, Cadence

October 20, 2011 Comments off

Long-term trends are strong for semiconductor and electronics. According to databeans estimate (Feb. 2011), semiconductor revenue will likely reach $450 billion by 2015 and electronics revenuw will likely reach $2,800 billion by 2015.

Speaking at the CDNLive! 2011 event in Bangalore, India, Charlie Huang, SVP of Worldwide Field Operations, Cadence Design Systems Inc., said that the challenges in the near term are slowdown in Europe and USA. The weakness is driven by increasingly negative views on the global economy, end demand, orders and outlook. Key indicators are also showing that the economy is facing headwinds. The 2011 GDP growth projections have deteriorated since the beginning of the year. The economy has been marred by high unemployment and low consumer confidence.

As of now, innovation has been driving growth. Apps have been driving innovation, followed by video, mobility, cloud and green technology. The impact on the electronics industry is multi-fold. There is a new development paradigm and collaboration has been increasing. The IP is also expanding beyond cores and the EDA is changing.

Source: Cadence.

Source: Cadence.

The new development paradigm for system companies is to differentiate on applications and semiconductor companies must deliver on application-driven hardware-software platforms. IP has now expanded well beyond the core. EDA is also changing, and Cadence is investing to deliver on the EDA360 vision. There are multiple silicon realization challenges. Cadence silicon realization solutions enable fast, deterministic, end-to-end path to silicon success.

As an example, ARM and Cadence have collaborated on the GHz implementation of Cortex-A15. ARM chose ARM Artisan physical IP, evaluated the Cortex-A15 RTL, and supported CPF. Cadence optimized the EDA flow, experienced support at EAC, and provided EDA tool releases and iRM.

ARM, TSMC and Cadence also collaborated on the industry’s first 20nm Cortex-A15. TSMC provided the 20nm process qualification and A15 learnings. ARM handled the 20nm implementation experience, A15 considerations in 20nm and TSMC 20nm readiness milestone. Cadence provided the 20nm research to reality, contributed and grew the A15 expertise and TSMC 20nm readiness milestone.

The end result: the industry’s first 20nm Cortex-A15 tapeout, thanks to a successful three-way vertical collaboration. ARM, Cadence and TSMC engineers worked side-by-side. The project priorities included 20nm DPT implementation schedule and 20nm readiness milestone.
Read more…

%d bloggers like this: