Semicon in sub-20nm era: Business as usual or different?

We are now entering the sub-20nm era. So, will it be business as usual or is it going to be different this time? With DAC 2013 around the corner, I met up with John Chilton, senior VP, Marketing and Strategic Development for Synopsys to find out more regarding the impact of new transistor structures on design and manufacturing, 450mm wafers and the impact of transistor variability.

Impact of new transistor structures on design and manufacturing
First, let us understand what will be the impact of new transistor structures on design and manufacturing.

John Chilton.

Chilton said: “Most of the impact is really on the manufacturing end since they are effectively 3D transistors. Traditional lithography methods would not work for manufacturing the tall and thin fins where self-aligned double patterning steps are now required.

“Our broad, production-proven products have all been updated to handle the complexity of FinFETs from both the manufacturing and the designer’s end.

“From the design implementation perspective, the foundries’ and Synopsys’ goal is to provide a transparent adoption process where the methodology (from Metal 1 and above) remains essentially the same as that of previous nodes where products have been updated to handle the process complexity.”

Given the scenario, will it be possible to introduce 450mm wafer handling and new lithography successfully?

According to Chilton: “This is a question best asked of the semiconductor manufacturers and equipment vendors. Our opinion is ‘very likely’.” The semiconductor manufacturers, equipment vendors, and the EDA tool providers have a long history of introducing new technology successfully when the economics of deploying the technology is favorable.

The 300nm wafer deployment was quite complex, but was completed, for example. The introduction of double patterning at 20nm is another recent example in which manufacturers, equipment vendors and EDA companies work together to deploy a new technology.

Impact of transistor variability and other physics issues
Finally, what will be the impact of transistor variability and other physics issues?

Chilton said that as transistor scaling progresses into FinFET technologies and beyond, the variability of device behavior becomes more prominent. There are several sources of device variability.

Random doping fluctuations (RDF) are a result of the statistical nature of the position and the discreteness of the electrical charge of the dopant atoms. Whereas in past technologies the effect of the dopant atoms could be treated as a continuum of charge, FinFETs are so small that the charge distribution of the dopant atoms becomes ‘lumpy’ and variable from one transistor to the next.

With the introduction of metal gates in the advanced CMOS processes, random work function fluctuations arising from the formation of finite-sized metal grains with different lattice orientations have also become important. In this effect, each metal grain in the gate, whose crystalline orientation is random, interacts with the underlying gate dielectric and silicon in a different way, with the consequence that the channel electrons no longer see a uniform gate potential.

The other key sources of variability are due to the random location of traps and the etching and lithography processes which produce slightly different dimensions in critical shapes such as fin width and gate length.

“The impact of these variability sources is evident in the output characteristics of FinFETs and circuits, and the systematic analysis of these effects has become a priority for technology development and IP design teams alike,” he added.

Convergence of PV materials, test and reliability: What really matters?

SEMI, USA recently hosted the seminar on ‘Convergence of PV Materials, Test and Reliability: What Really Matters?

Reliability in growing PV industry
Speaking on the importance of reliability to a growing PV industry, Sarah Kurtz, principal scientist, Reliability group manager, NREL, said that confidence in long-term performance is a necessity in the PV industry. Current failure rates are low. There is need to demonstrate confidence so that failure rates will stay low. There has been exponential growth of the PV industry so far. PV is a significant fraction of new installations. It now represents a significant fraction of new electricity generating installations of all kinds.

How does one predict the lifetime of PV modules? There has been a qualification test evolution for JPL block buys. Most studies of c-Si modules show module failures are small. Internal electrical current issues often dominate.

The vast majority of installations show very low PV module failure rates (often less than 0.1 percent). There has been evidence that PV is low risk compared to other investments. To sustain the current installation rate, we need to demonstrate confidence that justifies the annual investment of $100 million or so.

Critical factors in economic viability of PV
DuPont has broad capabilities under one roof. It offers materials, solar cell design, and processes integrated with panel engineering. Speaking about Critical factors in economic viability of PV – materials matter – Conrad Burke, global marketing director, DuPont PV Solutions, said that material suppliers have a distinct advantage to view trends. The industry can expect consolidation among large PV module producers and large materials suppliers.

There is an increasing dependence on materials suppliers for processes, tech support and roadmap. There is renewed attention to long-term reliability and quality of materials in PV products.

There is a race for survival among panel producers. There are dropping prices for solar panels, and quality is getting compromised. There are reduced incentives in established markets. The market will continue to grow. Key factors that determine investment return for PV include lifetime, efficiency and cost.

When materials fail, the consequences are dire. There are failures such as encapsulant discoloration, backsheet failure, glass delamination, etc. Average defect rates in new-build modules has been increasing. Significant number of PV installations do not deliver the projected RoI. The system lifetime is as important as cost and incentives.

Solar cell power continues to improve. There have been improvements from metal pastes and processes. Performance loss impacts the RoI. The US Department of Energy hired JPL to develop 30-year PV modules. Recent cost pressures have led to the dramatic changes in module materials and a lack of transparency.

Analyzing modules from the recent service environments show performance issues. Certification does not mitigate risk. Tests do not predict the actual field performance. He showed tier-1 solar panel manufacturing problems from China, Japan and the USA. Backsheet is critical to protect solar panels. Few materials have lengthy field experience. We will continue to see drop in prices for solar panels and opening of new markets. Focus for PV module makers will remain efficiency, etc.

Metal MEMS provides future of cell phone radio

Dennis Yost.

Cavendish Kinetics is well known for its combined experience in MEMS, RF system design and CMOS design. Since 2008, it has focused on developing digital variable capacitors to improve wireless connectivity and data rates for mobile phones.

According to Dennis Yost, president & CEO, Cavendish Kinetics, 4G/LTE mobile devices are not yet achieving their potential. Antenna frequency tuning is an essential technology. Only metal MEMS technology has the size and performance. He was speaking at the ongoing Globalpress Electronics Summit 2013 in Santa Cruz, USA.

Cavendish claims to have the team, proven technology and real demonstrated performance. There is IP and patent protection for customers. Cavendish also owns the process.

The future of cell phone radio is needed in order to meet the performance gap. In future, you will see adaptive power amplifiers.

Antenna frequency tuning used in traditional RF applications. How do you ensure there is no loss in the component? Only MEMS has the performance and size for cell phones. Metal MEMS has almost no series resistance. No switches are required.

Previous designs required switches and different loads. Mechanical capacitors change capacitance value by moving plates – changing the area or plate distance changes the capacitance. MEMS capacitors do the same at the micrometer level.

Successful MEMS
Users can control design and manufacturing process of devices. How a MEMS is built is just as important as what you build. Success requires MEMS design expertise, MEMS process expertise and MEMS volume production expertise.

Cavendish has MEMS experts in all areas. It developed and owned MEMS manufacturing process. It uses all standard CMOS foundry technology. Innovations have so far yielded over 100 patents in manufacturing process and MEMS design.

By using the NanoMech technology performance, Cavendish Kinectics has demonstrated excellent performance in a small chip.

Nanometer an enabling technology

Prof. Yi Cui

According to Prof. Yi Cui, Dept. of Materials, Science & Engineering, Stanford University, nanometer is an enabling technology. We can do applications such as electronics, energy, environment and health. Some examples are high energy batteries, printed energy storage devices on paper, textile and sponge, etc. He was delivering the inaugural address at the Globalpress Electronics Summit 2013, being held in Santa Cruz, USA.

High energy battery has portable and stationary applications. In portable, energy density, cost and safety are important. In stationary, cost, power, energy efficiency and ultra-long life are important. The standard is 500 cycles at 80 percent. One of the challenges of silicon anodes is that Si has 4200 mAh/g of silicon, 10 times more than carbon.

Nanowires can offer shorter distance for Li diffusion (high power), good strain release and interface control (for better cycle life), and continuous electron transport pathway (high power). In-situ transmission electron microscopy (TEM). Double walled hollow structure provides stable solid electrolyte interphase (SEI). The outer surface is static. Amprius, where Prof. Cui is CTO,  is a $6 million US government funded enterprise. Amprius China started in Nanjing, in April 2012.

Another example is printed energy storage devices on paper, textile and sponge. For low-cost scaffold, paper, textile and sponge, are used. There is cellulose paper and synthetic textile, besides sponge, as well.

There can be transparent batteries. It is actually very hard to develop those. The challenges for making a transparent battery are Al film, cathode, electrolyte, etc. An idea: dimension smaller than eye’s detection limit (50-100 um). Also, grids are well aligned.

Transparent conducting electrodes provide electrical and allow light to pass through. Apps include solar cells, etc. Indium tin oxide (ITO) has a low abundance of indium, brittleness when bent, and sputtering at high cost. Electrospinning of nanofibers is done for transparent electrodes. An example is the trough-shaped nanowires.

Yet another example is the water nanofilters for killing pathogens. The processes available for killing bacteria include chemical disinfection, UV disinfection, boiling, etc.

The first generation product is currently ready at Amprius. Amprius licensed the IP from Stanford. Stanford is also an investor in Amprius.

Five-year outlook for solar PV industry!

According to Finlay Coville, VP and team leader, NPD Solarbuzz, full year end market PV demand during 2012 reached 29.05 GW. The demand is forecast to increase to 31 GW in 2013. China is expected to replace Germany as the leading market for the first time. The global market is likely to have a CAGR exceeding 15 percent, highlighting long term confidence in global PV adoption levels.

Supply vs. demand overview in 2012
The upstream c-Si module/thin-film panel suppliers produced 30.1 GW of new product in 2012. Combined with inventory levels through the value chain, this provided 31 GW of panels to the downstream channels. 29 GW was used for market demand, while 2 GW went to the downstream inventory.

Demand overview 2013
Year 2013 is shaping up as a 31 GW demand year under the most likely scenario. Over 50 percent of the end market demand is projected to come from China, Germany and North America (USA and Canada). 2013 will be a transition year for the emerging PV territories. Both the Middle East and Africa and Emerging Asia will likely reach 1 GW.

Source: NPD Solarbuzz, USA.

PV demand in 2012 accounted for approximately 30 percent of all PV installed globally. The industry growth in 2012 is positive, but set against a backdrop of an industry that had been accustomed to year-on-year growth often exceeding 100 percent. The industry is forecast to return to double digit growth.

PV scenario forecasting continies to show divergent outcomes in 2017. A high market demand scenario assumes a strong economic environment and aggressive PV policies by way of direct incentives and lower regulatory hurdles.

Five-year cumulative demand by geography
Cumulatively, global PV demand is forecast to exceed 230 GW over the five year period to 2017. China is forecast to install 51 GW accounting for over 20 percent. Europe will continue to offer strong regional PV market. North America and Japan will provide over 61 GW of demand. Emerging markets are projected to create over 25 GW of PV demand, more than 10 percent of the cumulative total to 2017.

By application segment, the ground-mount segment will remain the single largest segment over the five years. Residential and non-residential (commercial) segments will continue to be characterized by specific end-user requirements, different supply channels and routes-to-market for upstream suppliers.

The PV industry was configured to supply over 45 GW in 2012. The industry is likely to be in an over-capacity mode in 2013, with balanced supply/demand levels restored from 2015. Market share aspirations remain a key driver for PV manufacturers. During 2013 and 2014, the capacity taken offline is likely to be more than compensated for by newly ramped capacity.

With multi-domain c-Si module production, most panels had efficiencies in the 13-16 percent band during 2012. High efficiency concepts are not likely to strongly influence the module efficiency landscape during 2013 or 2014. If high efficiency cell types gain traction, the share of modules with efficiencies above 16 percent will increase.

In 2012, a wide range of efficiencies were produced, but with levels that do not compete with c-Si modules for space-constrained applications. The range of panels available in the 12-14 percent band is likely to grow strongly from 2015 as leading suppliers benefit from process improvements. Panels below 10 percent efficiency will become obsolete.

Despite end market growth expected, revenues available to each part of the value-chain will see strong declines Y/Y in 2013. This is due to the ASPs declining at a faster rate than the end-market demand growth, within a strong overcapacity environment. Revenues are also unlikely to recover for each value-chain segment until the 2016-2017 period.

What’s with prices?
2012 was the fourth year in a row that c-Si module prices declined and was the largest Y/Y decline. As capacity throughout the PV chain has increased, the oversupply has put further pressure on the ASPs. Declines in pricing occurred further upstream, at the poly, wafer and cell segments.

Tracking SAM revenues fron selling modules into downstream channels is becoming less important to the PV industry. as a number of module suppliers take on EPC and project developer roles.

PV equipment spending
As for PV equipment spending, the most likely forecast sees capacity being added by a select gtoup of tier 1 c-Si makers during 2014. The next cyclic downturn is forecast for 2016-2017. This assumes excess capacity is added in the next upturn.

If we look at the current scope of trade disputes, there are five major markets — EU, USA, India, Canada, China — investigating products being imported, with China featuring in most cases. Most disputes are being pursued by the internal bodies, but several have been referred to the WTO for review. A growing number of emerging PV regions already have domestic content incentives.

Summary
PV demand was 29 GW in 2012, and 2013 is forecast to tip 31 GW. 230 GW of new PV demand is forecast between 2013-2017, adding to the 100 GW at the end of 2012. Eighty percent of PV demand in 2013-2017 will come from the top 10 end markets.

What technology SoC engineers need for next-gen chips?

About 318 engineers and managers completed a blind, anonymous survey on ‘On-Chip Communications Networks (OCCN), also referred to as an “on-chip networks”, defined as the entire interconnect fabric for an SoC. The on-chip communications network report was done by Sonics Inc. A summary of some of the highlights is as follows.

The average estimated time spent on designing, modifying and/or verifying on-chip communications networks was 28 percent (for the respondents that knew their estimate time).

The two biggest challenges for implementing OCCNs were meeting product specifications and balancing frequency, latency and throughput. Second tier challenges were integrating IP elements/sub-systems and getting timing closure.

As for 2013 SoC design expectations, a majority of respondents are targeting a core speed of at least 1 GHz for SoCs design starts within the next 12 months, based on those respondents that knew their target core speeds. Forty percent of respondents expect to have 2-5 power domain partitions for their next SoC design.

A variety of topologies are being considered for respondents’ next on-chip communications networks, including NoCs (half), followed by crossbars, multi-layer bus matrices and peripheral interconnects; respondents that knew their plans here, were seriously considering an average of 1.7 different topologies.

Source: Sonics Inc., USA.

Twenty percent of respondents stated they already had a commercial Network-on-Chip (NoC) implemented or plan to implement one in the next 12 months, while over a quarter plan to evaluate a NoC over the next 12 months. A NoC was defined as a configurable network interconnect that packetizes address/data for multicore SoCs.

For respondents who had an opinion when commercial Networks-on-Chip became an important consideration versus internal development when implementing an SoC, 43 percent said they would consider commercial NoCs at 10 or fewer cores; approximately two-thirds said they would consider commercial NoCs at 20 or fewer cores.

The survey participants’ top three criteria for selecting a Network on Chip were: scalability-adaptability, quality of service and system verification, followed by layout friendly, support for power domain partitioning. Half of respondents saw reduced wiring congestion as the primary reason to use virtual channels, followed by increased throughput and meeting system concurrency with limited bandwidth.

Components Direct offers guaranteed traceable E&O inventory!

Components Direct is a leading source for authorized end-of-life and excess electronic components. The products are guaranteed grade A factory sealed direct from the manufacturer and inventoried in a ESD 20.20 certified and ISO 9001 certified state-of-the art-facility. Components Direct is headquartered in Milpitas, CA with locations in the US and Asia.

It has a leading cloud-based platform for excess and obsolete (E&O) inventory. In 2012, Avnet and Components Direct entered in a strategic relationship. Components Direct is the exclusive channel for Avnet’s factory authorized excess and end-of-life components.

Compared to leading industry giants, such as Element14 and RS Components, Components Direct, currently, doesn’t have a detailed menu showcasing listed products, at least not on the home page, as yet. One hopes that’ll make an appearance soon.

Speaking on the mission of Components Direct, Anne Ting, executive VP, Marketing said: “Components Direct is the premier authorized distributor for excess and end-of-life electronic components. We are the only company working directly with manufacturers and their franchised distributors to offer 100 percent guaranteed traceable E&O components as well as technology services to combat counterfeit components and other gray market activity.

“For our supplier partners, we enable them to put excess product back into the control of an authorized source, as opposed to the gray market. For buyers, we provide them with a secure, authorized one-stop shop for excess, obsolete and unsold factory components.”

Combating gray market
How important is it to combat the gray market? Why will this endeavor stop/lessen gray market activity?

According to Ting, the gray market is a serious and growing problem. As early as 2008, a study by KPMG and the Alliance for Gray Market and Counterfeit Abatement (AGMA) stated that as much as $58 billion of technology products were passing through the gray market, and the problem has only gotten worse.

The gray market is rampant throughout all industries, with everyone from engineers, to procurement professionals and consumers impacted negatively when the products they purchase are advertised as new and authentic, but in reality could be used, refurbished or even worse, counterfeit.

In fact, a 2012 study by market research firm IHS found that over 12 million counterfeit electronics and semiconductor components

Anne Ting

have entered the distribution chain since 2007, with 57 percent of all counterfeit parts obsolete or end-of-life components. Many of these parts make their way into mission-critical industries, such as defense and aerospace, where a malfunctioning counterfeit part can mean the difference between life and death.

While provisions in the 2012 National Defense Authorization Act have enabled the government and trade groups to make some progress towards regulating the supply chain to ensure that components are only sourced directly from the manufacturers or their franchised distributors, the problem has not abated. The Act empowers the federal government to hold contractors financially responsible for replacing counterfeit products.

This, together with other changes, puts more responsibility on suppliers of electronic component to have risk mitigation procedures in place. The issue is become more topical and the industry must act in order to comply with the new legislation.

Components Direct takes this problem seriously, and provides supplier insights and tools to help combat gray market activity. In a recent study we conducted for a leading semiconductor supplier of both analog and digital devices, we discovered that over 124 million units of their product were floating in the gray market across 6,500 plus part numbers.

Over 70 percent of the products were found in Asia, and 20 percent also appeared in both North America and EMEA. The product age spanned many years with date codes of less than one year accounting for 22 percent of their gray market product. A further 5 percent had date codes over 11 years, demonstrating that whether you were an OEM looking for the newest product, or a military sub-contractor looking for obsolete components, no end customer is immune to the presence of unauthorized product.

Components Direct’s technology tools and services track gray market activity and provide suppliers with unprecedented visibility to their product leakage in the gray market by part number, region, data code etc. This data enables our suppliers to trace leakage in their supply chain and lessen potential unauthorized product from getting into the gray market.

Additionally, Components Direct provides suppliers and buyers with a secure, factory authorized channel for selling or purchasing 100 percent guaranteed traceable components. “We only sell products that come directly with manufacturers or their franchised distributors and all our products are inventoried in an ESD 20.20 and ISO 9001 certified facility,” said Ting.As an extension of the manufacturer, Components Direct provides the supply chain buyer with complete confidence and peace of mind that all products originate directly from the manufacturer and have been properly stored, handled and packaged. Sourcing from an authorized source like Components Direct eliminates the risks surrounding product quality, reliability and liability.

Unlocking full potential of soft IP

SoC design challenges and needs are diverse. There are many diverse IP blocks. It is time consuming to verify. Picking correct IP is critically important.

Speaking on the TSMC Soft IP Alliance program, Dan Kochpatcharin, deputy director, TSMC said that IP sourcing priorities include is it available, Is it from trusted partners, how is the design quality, and what are the specs and cost? Some other points to note are: is an IP verified, has it been silicon proven, what has been tested and how many in production volume already?

TSMC Soft IP Alliance program.

TSMC Soft IP Alliance has 5000+ IP titles from 40+ IP vendors. The IP Alliance program has been expanding. It is leveraging on successful IP. More and more customers are concerned about PPA data of soft IP specific technology when doing system design.

The Soft IP Alliance has 16 members. The Soft-IP quality assessment TSMC 9000 is key. New soft-IP handoff kit were rolled out in Nov. 2012. Major partners have now joined to drive soft-IP quality, such as Imagination, Sonics, MIPS, etc.

TSMC Soft-IP 9000 has carried out industry first QA assessment system for RTL based IP. TSMC and IP partners co-optimize RTL/process to deliver PPA optimized IPs.

Mike Gianfagna, Atrenta, spoke on the implementing program with Atrenta IP kit. Atrenta’s SpyGlass is a systematic approach to soft IP quality.

If you look at what’s needed for IP assessment, there are factors such as right abstraction levels must be supported, and soft IP is delivered as generators, RTL or gates; the biggest need is here. The IP must be comprehensive, easy to use, objective and quantifiable, actionable, and dynamic and scalable. Atrenta and TSMC announced SpyGlass IP kit 2.0 in October 2012

What does the IP kit check? Many items that would impact the integration/debug time and chip function were found and fixed. Soft IP qualification can be automated. It results in higher quality deliverables. All soft IP can be improved. Primary beneficiaries are chip integrators.

John Bainbridge, Sonics, spoke on the practical results of program participation. Sonics is a leader in system IP for SoCs. It enables designers to integrate any IP from anywhere, anytime.

Sonics helps leading SoC vendors solve some of the most difficult challenges in SoC design. These can be IP integration, high frequency, memory throughput, security, physical design, power management, development costs, and time-to-market. Sonics is a lead beta partner for TSMC Soft IP 2.0 kit program. It has worked closely with Atrenta and TSMC to ensure a seamless design flow.

What should India do to boost semiconductors?

I’ve already written a lot on the Indian semiconductor industry. Now, there’s nothing new to say. Even then, I am literally coaxed to say what I think the Indian semiconductor industry should do! As though the industry will listen to a nobody like me! 

First, the industry should stop wasting time running here and there, and focus on getting the job done! Semiconductors isn’t a new area, and has been in existence even before the India Semiconductor Association (ISA) came into being in 2005.

There have been talks (ongoing since about 2006) about building fabs in India. Well, where are they? Back in 2010, I wrote a post titled Indian industry proposes to extend deadline of India’s semicon policy up to March 2015! One sincerely hopes that has actually happened!

India could consider building 150/180/200mm fabs that tackle local problems via indigenous applications. And, there are scores of local issues that need to be dealt with! I’ve said before, and am repeating myself at the sake of repetition — the semiconductor industry is NOT the IT industry, but it appears to being treated like one, especially in India!

Indian companies could consider developing firms in the assembly testing, verification and packaging (ATMP) space. Very little has happened so far and a lot more needs to be done. There could be some attempts to attract and invite companies in areas such as RFID to address local problems and develop local applications, unless India has given up on RFIDs.

I really have very little idea whether there is any interest in India to pursue global companies in PDP, OLED/LED space for setting up manufacturing units. Although, I can safely bet that if it is the Chinese companies that Indian firms are setting themselves up to take on, we would have a very long way to go!

India also needs to kindly forget about the ‘states race’! It has not helped anyone so far, nor will it help anyone in future!! In the end, we are all looking to develop India, aren’t we?

I didn’t even know that there is so much time required for setting up a pan-industry panel that will determine the top five products that are important for India! Seriously!! Anyone, who resides in India, should be able to tell you that the key sectors in India are automotive, consumer, industrial, medical and telecom. Agree that automotive and certain medical electronics areas can be expensive. Well, there are still three areas to pursue!

If anyone had simply bothered to send me an email or even call me, I’d have very happily told them about the top five product lines that are important for India and much more!   There is a pressing need to develop a robust Indian semiconductor industry, led by local companies! Many would agree that all of this seems very easy to say, but difficult to manage!