Part of the  

Chip Design Magazine

  Network

About  |  Contact

Posts Tagged ‘China’

China’s Bold Strategy for Semiconductors

Thursday, October 20th, 2016

Gabe Moretti, Senior Editor

The East-West Center is a research organization Established by the U.S. Congress in 1960. The Center serves as a resource for information and analysis on critical issues of common concern, bringing people together to exchange views, build expertise, and develop policy options. The Center is an independent, public, nonprofit organization with funding from the U.S. government, and additional support provided by private agencies, individuals, foundations, corporations, and governments in the region.

The Center’s 21-acre Honolulu campus, adjacent to the University of Hawai‘i at Mānoa, is located midway between Asia and the U.S. mainland and features research, residential, and international conference facilities.  A few years ago I became acquainted with Dr. Dieter Ernst, a Senior Fellow at the Center.  He has recently published a paper with the title: ”China Bold Strategy for Semiconductors – Zero Sum Game or Catalyst for Cooperation?”

Abstract of the Paper

This paper explores whether China’s bold strategy for semiconductors will give rise to a zero-sum game or whether it will enhance cooperation that will benefit from increased innovation in China.  As the world’s largest producer and exporter of electronic products, China is by far the top market for integrated circuits (ICs), accounting for nearly a third of global demand. Yet its ability to design and produce this critical input remains seriously constrained. Despite decades and many billions of dollars of state-led investment, China’s domestic production of semiconductors covers less than 13% of the country’s demand.

As a result, China’s IC trade deficit has more than doubled since 2005, and now has surpassed crude oil to become China’s biggest import item. To correct this unsustainable imbalance, China seeks to move from catching up to forging ahead in semiconductors through progressive import substitution. The “National Semiconductor Industry Development Guidelines (Guidelines)” and the ”Made in China 2025″ (MIC 2025) plan were published by China’s State Council in June 2014 and May 2015, respectively. Both plans are backed by huge investments and a range of support policies covering intellectual property, cybersecurity, procurement, standards, rules of competition (through the “Anti-Monopoly Law”), and the negotiation of trade agreements, like the Information Technology Agreement. The objective is to strengthen simultaneously advanced manufacturing, product development and innovation capabilities in China’s semiconductor industry as well as in strategic industries that are heavy consumers of semiconductors.

Until recently, China has focused primarily on logic semiconductors and mixed-signal integrated circuits for mobile communication equipment (including smart phones), and on the assembly, testing and packaging of chips. Since the start of the 13th FYP, China’s semiconductor industry strategy now covers a much broader range of products and value chain stages, while at the same time increasing the depth and sophistication of its industrial upgrading efforts.

Based on a review of policy documents and interviews with China-based industry experts, Dr. Ernst describes a key policy initiatives and stakeholders involved in the current strategy; highlight important recent adjustments in the strategy to broaden China’s semiconductor product mix; and assess the potential for success of China’s ambitious efforts to diversify into memory semiconductors, analog semiconductors, and new semiconductor materials (compound semiconductors). The chances for success are real, giving rise to widespread worries in the US and across Asia that China’s bold strategy for semiconductors may result in a zero-sum game with disruptive effects on markets and value chains. However, Chinese semiconductor firms still have a long way to go to catch up with global industry leaders. Hence, global cooperation to integrate China into the semiconductor value chain makes more sense than ever, both for the incumbents and for China.

More About the Plan

Dr. Ernst goes to great details in his paper to describe the latest Chinese effort in semiconductors.  To begin with the present leadership team includes, contrary to the past, internationally recognized scientists and technical leaders.  The effort is focused on few areas of the industry and seems well managed.  One focus area is the design and fabrication of power and analog semiconductors especially with regards to the requirements for robotic applications.  In the paper Dr. Ernst writes: “On the demand side, China’s well funded programs to develop both electric vehicles and smart autonomous buses and cars will create a huge demand for analog semiconductors.”  Other areas that need analog devices are: smart grid, alternative energy technologies, and IoT systems.

On the supply side, Dr. Ernst points out, “analog semiconductors offer substantial advantages – they use mature process technologies, and thus are much more cost effective than digital fabs.”  This and other related advantages over digital IC design and fabrication make the choice an intelligent one especially manufacturing costs.

Dr. Ernst states that: “Of particular interest will be China’s push into compound semiconductors.  While still at an early stage, there are serious efforts under way to develop an integrated compound semiconductor value chain, drawing on the demand pull from China’s huge market for lighting/LED and power electronics.”  The paper details the names of companies, not all Chinese by the way, that are part of the effort.

Memory is a new sector of interest to the Chinese government.  In the past this segment of the industry had been neglected, but the new plan is now considering it important with significant investment for both flash memory and DRAM products.

In short, the present Chinese plan is very serious, focused, and so far, well managed.  China in a few years could become a serious disruptor of present semiconductor commerce.  American companies, as well as Taiwanese, Japanese, and South Korean, need to pay particular attention to Chinese efforts in semiconductors.   China could not only cover most of its internal needs, but can in fact develop into an international exporter of ICs.

Conclusion

Dr. Ernst paper goes into great details about the Chinese strategy for semiconductors.  What I have done is just provide highlights.  I strongly believe that the paper is must read for all those in the EDA, systems, and foundry business.  Not just to follow what the Chinese government is doing, but also to extract possible ideas on what the US companies might need to do to maintain their commercial and technological lead.  The entire paper can be found at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2836331.

The EDA Industry Macro Projections for 2016

Monday, January 25th, 2016

Gabe Moretti, Senior Editor

How the EDA industry will fare in 2016 will be influenced by the worldwide financial climate. Instability in oil prices, the Middle East wars and the unpredictability of the Chinese market will indirectly influence the EDA industry.  EDA has seen significant growth since 1996, but the growth is indirectly influenced by the overall health of the financial community (see Figure 1).

Figure 1. EDA Quarterly Revenue Report from EDA Consortium

China has been a growing market for EDA tools and Chinese consumers have purchased a significant number of semiconductors based products in the recent past.  Consumer products demand is slowing, and China’s financial health is being questioned.  The result is that demand for EDA tools may be less than in 2015.   I have received so many forecasts for 2016 that I have decided to brake the subject into two articles.  The first article will cover the macro aspects, while the second will focus more on specific tools and market segments.

Economy and Technology

EDA itself is changing.  Here is what Bob Smith, executive director of the EDA consortium has to say:

“Cooperation and competition will be the watchwords for 2016 in our industry. The ecosystem and all the players are responsible for driving designs into the semiconductor manufacturing ecosystem. Success is highly dependent on traditional EDA, but we are realizing that there are many other critical components, including semiconductor IP, embedded software and advanced packaging such as 3D-IC. In other words, our industry is a “design ecosystem” feeding the manufacturing sector. The various players in our ecosystem are realizing that we can and should work together to increase the collective growth of our industry. Expect to see industry organizations serving as the intermediaries to bring these various constituents together.”

Bob Smith’s words acknowledge that the term “system” has taken a new meaning in EDA.  We are no longer talking about developing a hardware system, or even a hardware/software system.  A system today includes digital and analog hardware, software both at the system and application level, MEMS, third party IP, and connectivity and co-execution with other systems.  EDA vendors are morphing in order to accommodate these new requirements.  Change is difficult because it implies error as well as successes, and 2016 will be a year of changes.

Lucio Lanza, managing director of Lanza techVentures and a recipient of the Phil Kaufman award, describes it this way:

“We’ve gone from computers talking to each other to an era of PCs connecting people using PCs. Today, the connections of people and devices seem irrelevant. As we move to the Internet of Things, things will get connected to other things and won’t go through people. In fact, I call it the World of Things not IoT and the implications are vast for EDA, the semiconductor industry and society. The EDA community has been the enabler for this connected phenomenon. We now have a rare opportunity to be more creative in our thinking about where the technology is going and how we can assist in getting there in a positive and meaningful way.”

Ranjit Adhikary, director of Marketing at Cliosoft acknowledges the growing need for tools integration in his remarks:

“The world is currently undergoing a quiet revolution akin to the dot com boom in the late 1990s. There has been a growing effort to slowly but surely provide connectivity between various physical objects and enable them to share and exchange data and manage the devices using smartphones. The labors of these efforts have started to bear fruit and we can see that in the automotive and consumables industries. What this implies from a semiconductor standpoint is that the number of shipments of analog and RF ICs will grow at a remarkable pace and there will be increased efforts from design companies to have digital, analog and RF components in the same SoC. From an EDA standpoint, different players will also collaborate to share the same databases. An example of this would be Keysight Technologies and Cadence Designs Systems on OpenAccess libraries. Design companies will seek to improve the design methodologies and increase the use of IPs to ensure a faster turnaround time for SoCs. From an infrastructure standpoint a growing number of design companies will invest more in the design data and IP management to ensure better design collaboration between design teams located at geographically dispersed locations as well as to maximize their resources.”

Michiel Ligthart, president and chief operating officer at Verific Design Automation points to the need to integrate tools from various sources to achieve the most effective design flow:

“One of the more interesting trends Verific has observed over the last five years is the differentiation strategy adopted by a variety of large and small CAD departments. Single-vendor tool flows do not meet all requirements. Instead, IDMs outline their needs and devise their own design and verification flow to improve over their competition. That trend will only become more pronounced in 2016.”

New and Expanding Markets

The focus toward IoT applications has opened up new markets as well as expanded existing ones.  For example the automotive market is looking to new functionalities both in car and car-to-car applications.

Raik Brinkmann, president and chief executive officer at OneSpin Solutions wrote:

“OneSpin Solutions has witnessed the push toward automotive safety for more than two years. Demand will further increase as designers learn how to apply the ISO26262 standard. I’m not sure that security will come to the forefront in 2016 because there no standards as yet and ad hoc approaches will dominate. However, the pressure for security standards will be high, just as ISO26262 was for automotive.”

Michael Buehler-Garcia, Mentor Graphics Calibre Design Solutions, Senior Director of Marketing notes that many of the established and thought of as obsolete process nodes will instead see increased volume due to the technologies required to implement IoT architectures.

“As cutting-edge process nodes entail ever higher non-recurring engineering (NRE) costs, ‘More than Moore’ technologies are moving from the “press release” stage to broader adoption. One consequence of this adoption has been a renewed interest in more established processes. Historical older process node users, such as analog design, RFCMOS, and microelectromechanical systems (MEMS), are now being joined by silicon photonics, standalone radios, and standalone memory controllers as part of a 3D-IC implementation. In addition, the Internet of Things (IoT) functionality we crave is being driven by a “milli-cents for nano-acres of silicon,” which aligns with the increase in designs targeted for established nodes (130 nm and older). New physical verification techniques developed for advanced nodes can simplify life for design companies working at established nodes by reducing the dependency on human intervention. In 2016, we expect to see more adoption of advanced software solutions such as reliability checking, pattern matching, “smart” fill, advanced extraction solutions, “chip out” package assembly verification, and waiver processing to help IC designers implement more complex designs on established nodes. We also foresee this renewed interest in established nodes driving tighter capacity access, which in turn will drive increased use of design optimization techniques, such as DFM scoring, filling analysis, and critical area analysis, to help maximize the robustness of designs in established nodes.”

Warren Kurisu, Director of Product Management, Mentor Graphics Embedded Systems Division points to wearables, another sector within the IoT market, as an opportunity for expansion.

“We are seeing multiple trends. Wearables are increasing in functionality and complexity enabled by the availability of advanced low-power heterogeneous multicore architectures and the availability of power management tools. The IoT continues to gain momentum as we are now seeing a heavier demand for intelligent, customizable IoT gateways. Further, the emergence of IoT 2.0 has placed a new emphasis on end-to-end security from the cloud and gateway right down to the edge device.”

Power management is one of the areas that has seen significant concentration on the part of EDA vendors.  But not much has been said about battery technology.  Shreefal Mehta, president and CEO of Paper Battery Company offered the following observations.

“The year 2016 will be the year we see tremendous advances in energy storage and management.   The gap between the rate of growth of our electronic devices and the battery energy that fuels them will increase to a tipping point.   On average, battery energy density has only grown 12% while electronic capabilities have more than doubled annually.  The need for increased energy and power density will be a major trend in 2016.  More energy-efficient processors and sensors will be deployed into the market, requiring smaller, safer, longer-lasting and higher-performing energy sources. Today’s batteries won’t cut it.

Wireless devices and sensors that need pulses of peak power to transmit compute and/or perform analog functions will continue to create a tension between the need for peak power pulses and long energy cycles. For example, cell phone transmission and Bluetooth peripherals are, as a whole, low power but the peak power requirements are several orders of magnitude greater than the average power consumption.  Hence, new, hybrid power solutions will begin to emerge especially where energy-efficient delivery is needed with peak power and as the ratio of average to peak grows significantly. 

Traditional batteries will continue to improve in offering higher energy at lower prices, but current lithium ion will reach a limit in the balance between energy and power in a single cell with new materials and nanostructure electrodes being needed to provide high power and energy.  This situation is aggravated by the push towards physically smaller form factors where energy and power densities diverge significantly. Current efforts in various companies and universities are promising but will take a few more years to bring to market.

The Supercapacitor market is poised for growth in 2016 with an expected CAGR of 19% through 2020.  Between the need for more efficient form factors, high energy density and peak power performance, a new form of supercapacitors will power the ever increasing demands of portable electronics. The Hybrid supercapacitor is the bridge between the high energy batteries and high power supercapacitors. Because these devices are higher energy than traditional supercapacitors and higher power than batteries they may either be used in conjunction with or completely replace battery systems. Due to the way we are using our smartphones, supercapacitors will find a good use model there as well as applications ranging from transportation to enterprise storage.

Memory in smartphones and tablets containing solid state drives (SSDs) will become more and more accustomed to architectures which manage non-volatile cache in a manner which preserves content in the event of power failure. These devices will use large swaths of video and the media data will be stored on RAM (backed with FLASH) which can allow frequent overwrites in these mobile devices without the wear-out degradation that would significantly reduce the life of the FLASH memory if used for all storage. To meet the data integrity concerns of this shadowed memory, supercapacitors will take a prominent role in supplying bridge power in the event of an energy-depleted battery, thereby adding significant value and performance to mobile entertainment and computing devices.

Finally, safety issues with lithium ion batteries have just become front and center and will continue to plague the industry and manufacturing environments.  Flaming hoverboards, shipment and air travel restrictions on lithium batteries render the future of personal battery power questionable. Improved testing and more regulations will come to pass, however because of the widespread use of battery-powered devices safety will become a key factor.   What we will see in 2016 is the emergence of the hybrid supercapacitor, which offers a high-capacity alternative to Lithium batteries in terms of power efficiency. This alternative can operate over a wide temperature range, have long cycle lives and – most importantly are safe. “

Greg Schmergel, CEO, Founder and President of memory-maker Nantero, Inc points out that just as new power storage devices will open new opportunities so will new memory devices.

“With the traditional memories, DRAM and flash, nearing the end of the scaling roadmap, new memories will emerge and change memory from a standard commodity to a potentially powerful competitive advantage.  As an example, NRAM products such as multi-GB high-speed DDR4-compatible nonvolatile standalone memories are already being designed, giving new options to designers who can take advantage of the combination of nonvolatility, high speed, high density and low power.  The emergence of next-generation nonvolatile memory which is faster than flash will enable new and creative systems architectures to be created which will provide substantial customer value.”

Jin Zhang, Vice President of Marketing and Customer Relations at Oski Technology is of the opinion that the formal methods sector is an excellent prospect to increase the EDA market.

“Formal verification adoption is growing rapidly worldwide and that will continue into 2016. Not surprisingly, the U.S. market leads the way, with China following a close second. Usage is especially apparent in China where a heavy investment has been made in the semiconductor industry, particularly in CPU designs. Many companies are starting to build internal formal groups. Chinese project teams are discovering the benefits of improving design qualities using Formal Sign-off Methodology.”

These market forces are fueling the growth of specific design areas that are supported by EDA tools.  In the companion article some of these areas will be discussed.

DVCon Europe Is The Wrong Venue

Thursday, March 13th, 2014

Gabe Moretti, Contributing Editor

DVCon USA, or should I call it DVCon SV for Silicon Valley, has just concluded another successful year.  The conference has grown significantly, and it compares favorably with other SV based conferences in our industry.  It is now a must attend event for all professionals involved in design verification.  On the heels of such growth Accellera Systems Initiative has announced the first European DVCon conference to be held this October 14 – 15 in Munich Germany.  I believe this is a suboptimal choice.

<b>Toward The Past, Not The Future</b>

Discussions among highly placed members of the Accellera board about starting a conference outside of the US have been going on for at least a couple of years, but the location most talked about was India, not Europe.

In spite of the attraction of Munich in October as a tourist destination, I have reservations as to the wisdom of the choice.  Europe already has a number of EDA and electronics conferences, and one more is not going to make much difference.   Europe is a mature market for the three key corporate powers within Accellera: Mentor, Synopsys, and Cadence. It is not clear what they expect to gain from the conference.

Accellera Technical Working Groups already enjoy good participation from European technologists.  What is needed is to involve representatives from the growth markets.  We need to know what they need, we already know what the US and European users need or think they need.

It would make much more sense to continue to pursue the India idea, although a modification of the venue to China instead of India would make sense.  I agree that the Europe location is easier to arrange.  There are European corporate members in Accellera who can offer support and organizational knowhow, but where is the drive to expand, both technically and economically?

I anticipate rebuttals from some of the board members pointing out that they have a working relationship with the ASP-DAC conference in Japan and thus Asia is serviced, albeit indirectly..  But Japan suffers from the same characteristics as Europe.  Japanese technologists already work within the Accellera Technical Working Groups, and Japan is a mature market for EDA companies.

Growth needs new opportunities and new ideas, not recognition of past and present contributions, no matter how substantial and noteworthy they might have been.  So Accellera board members enjoy your liter of new beer this fall, but think about how many new corporate and technical WG members you could get from locating a conference in China.