By: Gary Dagastine

GLOBALFOUNDRIES was recently ranked No. 10 on SIA’s list of the top U.S. corporate patent recipients for 2016. We’re proud of that ranking because it’s more than just a number, it’s proof that the contributions GF employees make every day really matter, both for our company and for the industry overall.

We wanted to understand what it takes to make such an achievement possible because technology development is at the heart what we do. GF is full of smart, talented people but of course we couldn’t talk with everyone, so instead we contacted some of GF’s top patent recipients to get their insights. Each is a GLOBALFOUNDRIES Master Inventor.

Mukta Farooq is the Team Leader for 7nm Chip-Package Interaction and Packaging Technology. A GLOBALFOUNDRIES Fellow with 190 issued patents, she also is an IEEE Fellow, and a Distinguished Lecturer of the IEEE Electron Devices Society. She was previously Lifetime Master Inventor and Member of the Academy of Technology at IBM. Her areas of focus encompass 2D, 2.5D, and 3D chip package interaction and interconnect technologies.

Anthony (Tony) Stamper is a Distinguished Member of the Technical Staff in Analog and Mixed-Signal Technology Development in Essex Junction. Also previously a Lifetime Master Inventor at IBM with more than 400 issued patents, his work has spanned pioneering processes and integration technologies such as low-temperature CVD dielectric development, low-k dielectrics, damascene copper wire integration and, more recently, state-of-the-art RF device design, process integration and program management.

Ruilong Xie is a Senior Member of the Technical Staff in Albany, working on advanced FinFET technology as part of the IBM Technology Development Alliance, where he is the lead FEOL integrator at the 5nm node. An IEEE Senior Member, he has more than 200 issued patents for his work with FinFETs and for other device types such as gate-all-around and vertical FETs. His published work has been cited more than 600 times worldwide.

GF: How do you look at the process of technology development?

Tony: The closer one is to state-of-the-art manufacturing development, the easier it is to exceed the thresholds for patent novelty and patent value. My work has focused on differentiated technologies that can lead to industry “firsts” and the generation of intellectual property and many patents that have great value. By differentiated technologies I don’t only mean simply trying to develop the fastest transistor, but also the development of new IC technologies for wiring performance, cost, yield, reliability, etc. – all the things which differentiate us from competitors and lead to success in the marketplace.

Mukta: Working in new areas of technology often means that things don’t work out initially, and that many attempts may be needed before something clicks. A case in point is 3D technology where copper Through-Silicon-Vias (TSVs) were integrated into 32 nm CMOS logic wafers. This endeavor had significant challenges early on. But, as we got deeper into the technology and gained understanding of the problems, we began to invent original, creative ways of getting around them, leading not only to the industry’s first 3D Cu TSV logic wafer, but also to intellectual property associated with it!

GF: In a company made up of incredibly smart and dedicated people, you are a leader in terms of issued patents. What has made this performance possible, and is there any advice you would like to give to others?

Ruilong: I’d like to thank all my managers and mentors who gave me the opportunity early on to work on almost all FEOL modules, which in turn enabled me to work with, and learn from, all module owners and other technology teams. These interactions helped me develop a good understanding of the big picture and also built a natural collaboration with many brilliant and talented people.

On a more personal level, a problem-solving frame of mind and persistence in thinking are essential. A positive attitude towards problems can help generating the creativity needed to solve the issues and problems we face. Very often, the best solution does not just come out immediately. Persistence in thinking, even in the background mode, allows the brain to process all related information and do divergent thinking. This usually leads to very good solutions after an incubation period.

Mukta: The advice I would give is that first you need to develop a level of expertise in a given field and become familiar with the existing art. After you gain familiarity, you can think of how to improve on existing structures and methods to deliver benefits. Once you have reached that level of knowledge, you can start inventing. Also, collaborating with others – especially those who may have different backgrounds – can result in unique approaches to solving technology problems.

Tony: Persistence, hard work, and teamwork. There is a mantra at work that there are no problems, just opportunities. From a patent perspective, problems are opportunities to author patents. Roughly half of the patents I have authored stemmed from the identification and solution of technical problems. Most of the patents I authored had multiple authors. Working in informal or formal patent development teams is invaluable.

GF: What does GLOBALFOUNDRIES do to encourage the pursuit of technical achievement?

Mukta: The race to get to the best technological solution is ever-present in semiconductor companies. While technologists will do the best they can, recognizing and rewarding innovation encourages out-of-the-box thinking. GF provides this encouragement with a rich patent awards program, and with the Master Inventor Program that serves to highlight and appreciate the innovators for their contributions to the company’s intellectual property.

Tony: GF encourages patent generation through its patent review boards and financial incentives. Having an algorithmic approach to fostering specific inventions helps foster innovation overall. In technology development, there typically are core teams of a few engineers focusing on certain problem areas. These core teams often set aside time regularly to meet and identify patent-able ideas. This openness to brainstorm and write patents fosters a creative environment at GLOBALFOUNDRIES.

Ruilong: GF has a learning-oriented culture which encourages new ideas and experimentation. Jobs are highly aligned with employee competencies, and usually there is a high degree of autonomy. Also, open communication among leaders and coworkers leads to mutual support. Here’s a recent example: In our 7nm development work, we encountered a big challenge in a module called the “gate cut.” To resolve the issue, we encouraged ideas from all resources. Very effective discussions were held between Albany and Malta, with experts from both sides who met, shared past learnings, and brainstormed together. In the end, almost every possible solution was generated. We carefully reviewed each one, ranked and prioritized the top ones to pursue in experiments, and several high-quality patents have also been filed.

GF: Thank you for sharing your thoughts and insights!

 

During an era when semiconductor manufacturing was shifting to Asia, GF chose to fight the tide and invest in America, building a fab in upstate New York. New York State was keen to participate, as they had already invested in the region with a vision of creating a new technology ecosystem as an engine for job creation. These investments have been critical to GF’s success.

Dan Hutcheson, a leading semiconductor industry analyst and CEO of VLSI Research, recently visited Fab 8 in Saratoga County, New York and spent a day touring the facility and engaging with various fab teams.  The day concluded with a conversation between Dan and Tom Caulfield, senior vice president and general manager of Fab 8.

In this interview, Dan examines the accomplishments of GF’s Fab 8—where first-time-right on 14nm semiconductor design tapeouts has become a rule of thumb—and the importance of human capital, the impact of teamwork, and how to inspire it are essential components for success. They also uncover what semiconductor fab management is all about, including the difficulties of bringing up a greenfield fab, how the IBM Microelectronics acquisition fits in, and the role that Sanjay Jha, GF’s CEO, has played in the turnaround.

In short, Dan sums it up best: “Fab 8 is not just another fab. It is an American manufacturing success story.”

By: Igor Arsovski

GLOBALFOUNDRIES’ Ternary Content Addressable Memory Core sets a new performance and density record enabling faster cloud and data center traffic.

Data centers, which are the engines of the Internet, are responsible for the vast amount of traffic that flows across the network. And, it’s no revelation; our digital universe and data center traffic will reach 10.4 zettabytes (ZB) by 2019, the equivalent to 144 trillion hours of streaming music. As apps and services become more data hungry, the higher the allowance for data traffic we’re going to need to feed them.

And, getting there will require innovative solutions – to improve scalability and achieve higher throughput with lower latency, data centers use Ternary Content Addressable memory (TCAM).  TCAM, an application specific memory, is designed to help data centers speed the search of large look-up tables.

GF recently introduced a new ASIC search architecture TCAM solution at the International Solid-State Circuits Conference in San Francisco. GF’s TCAM core demonstrated the highest published search rate of 1.4 billion searches per second core, while also achieving a record density of 2Mb/mm2. This is good news for data centers, which commonly use TCAM in applications such as pattern recognition, data compression, network security and packet forwarding.

Furthermore, GF’s TCAM design uses both proprietary architecture and circuits for power supply pre-conditioning that reduce power supply noise collapse by 50 percent, enabling a path to further performance and density scaling. These improvements are critical to enable faster data centers with larger network tables capable of handling the constantly increasing network traffic.

TCAMs are extremely complex memory designs, which is why GF’s expertise – more than 20 years of pushing the limits of TCAM design – has enabled us to deliver improved search throughput, high density, and low power consumption for our customers’ ASIC designs. Specifically, it may be our seven generations of embedded TCAM design, qualification and debug experience, with more than 50 ASICs with large TCAM content in production, that allowed us to provide the highest search-bandwidth (fastest) networking TCAM in the industry.

To learn more about GF’s ASIC solutions, download our ASIC FX-14 technology brief or contact a GF ASIC sales representative.

 

Our 300mm Fab 1 is in Dresden, in the region of Germany known as Saxony, or sometimes Old Saxony. It is quite a beautiful and tranquil area but don’t be fooled. Years ago Saxony was Germany’s leading industrial area, and more recently it is again achieving industrial prominence. GLOBALFOUNDRIES is one of the reasons why. Dresden is where we are ramping 22FDX®, our 22nm FD-SOI process, into volume production this year. 22FDX gives customers an exceptional balance of performance, power consumption and cost, making it ideal for mobile, automotive and IoT applications. I’ve been managing Fab 1 in Dresden since 2011 and the differences from then to now can hardly be overstated. When I got here our production capabilities, technical resources and mode of operation didn’t match our ambition. Today, Fab 1 is quite simply one of the best fabs in the entire world, and we have a superb staff of technical people, managers and operating personnel who are aligned on achieving world-class operational metrics and customer satisfaction.

GLOBALFOUNDRIES, Fab 1, in Dresden is Europe’s Largest 300mm Factory

This evolution has been driven by investment, by more efficient use of technology and by great execution. We have invested more than $12 billion in Fab 1 to date, giving us the capacity to ship 60,000 wafers per month. This quarter, Dresden shipped its one millionth wafer to a top-tier mobile customer. That is an outstanding achievement. Overall, we have shipped 2.8 million wafers from Fab 1 since 2010, the majority of which were fabricated with our advanced HKMG technology. We’ve been offering that technology longer than any other pure-play foundry. We are currently making further investments to grow production capacity by another 40% by 2020. Much of this investment is to support volume production of 22FDX, along with the next-generation 12FDX™ technology now under development. Customer tape-outs from the 12FDX process are expected by the  end of 2018. Our efficient use of technology is a great advantage when it comes to volume FDX™ production. More than 70% of the FDX process steps are copied from our already well-established 28nm HKMG process flow, making it easier, faster and less risky to achieve commercial production. Being a foundry, we understand that we have one more key item on our daily agenda (besides shipping wafers on time and in spec, as well as developing tomorrow’s technologies): enhance productivity while reducing cost. That is now part of our site DNA, and I am proud that in a site-wide effort we were able to take out $100 million (USD) cost in 2016 alone.

GLOBALFOUNDRIES: Clean room techs walk the bridge in Fab 1

With regard to execution, over the last five years we have placed a great focus on establishing the necessary systems, business processes and mindset to complement our state-of-the-art manufacturing plant. We have learned how to better control risk and costs while making process changes, which is critical to managing the ever-increasing complexity that comes with more customers, more products, and new processes such as FDX. It’s been an exciting journey, and the best is yet to come.

In recent days, we have seen the world’s two largest chip manufacturers announce new low-power 22nm process technologies. We are pleased to see other manufacturers following our technology solution lead. Almost two years ago, we launched our 22FDX technology for wireless, battery-powered intelligent systems.

While others have been focused only on squeezing more digital performance out of the bleeding edge, we at GF have been focusing on system-level metrics with 22FDX. The 22nm node is now becoming one of the biggest battlegrounds in semiconductors, which demonstrates the unprecedented innovation that is taking place at advanced nodes that are one or two steps off of the leading edge.

We chose FD-SOI over bulk planar or FinFET because it offers the best combination of performance, power and area for these applications. 22FDX is unmatched by any competing 22nm technology in several areas:

• The lowest operating voltage (0.4 volt)
• The most cost-effective CPP and MX scaling
• The lowest mask count
• The best RF performance
• The only technology to offer software-controlled transistor body-biasing

And while others are announcing their offerings now, 22FDX is fully qualified for production at our Fab 1 facility in Dresden, Germany. We are seeing strong customer demand, with more than 50 active engagements in high-growth areas such as mobile, IoT, and automotive.

Not only are we delivering 22FDX now, but we are investing for the future. In September of last year, we introduced an extension to our FDX roadmap with our next generation, the 12FDX platform, which is the only technology to provide the design flexibility and cost of a planar process down to 12nm, and yet offer 10nm FinFET-level performance. We expect others to follow our 12FDX lead!

We are also making significant capacity investments. In Germany, we are building up 22FDX capacity, with plans to grow the overall fab capacity by 40 percent by 2020. And in China, we recently announced a partnership to build a 300mm fab in Chengdu to support the growth of the Chinese semiconductor market and to meet accelerating global customer demand for 22FDX.

Last but not least, a robust design and IP ecosystem is coalescing around FDX. We recently launched our FDXcelerator™ Partner Program, which is designed to reduce time-to-market for customers and facilitate faster migration to FD-SOI from bulk nodes such as 40nm and 28nm. More than 20 partners have joined the ecosystem since its launch only six months ago.

Clearly excitement about FDX is rising and the competition is heating up. FD-SOI has long been considered the “technology of the future,” but now all of the puzzle pieces are coming together to make the promise of FD-SOI a reality TODAY.