Parisians believes the Isle de France is the fountainhead of all things French. When one arrives in Nice, however, it is clear that the Cote d'Azure did not get the memo. Silicon Valley believes it is the fountainhead of all things EDA. When one arrives at DATE, however, it is clear that the European EDA community did not get the memo.
This past week, Nice felt like the center of all things French; DATE felt like the center of all things EDA. Sorry to disappoint, but Nice and DATE may be right. Nice was lively, relaxed, and colorful; DATE'09 was lively, energetic, and full of a variety of languages and viewpoints - most being system-level, peppered with multicore and a dollop of governmental instigation thrown in for good measure.
Europe is not North America, and vice-versa. North America, Silicon Valley in particular, prides itself on pulling progress, through entrepreneuralism and innovation; Europe is a lot more about pushing through governmental investments and societal visioning. In truth, neither geography is all pull or all push. They are both, both. It's simply a matter of degree.
This is my fifth visit to DATE. Each time I am re-energized by the enthusiasm and innovative twist the European community brings to the engineering and business of design automation. Following are snapshots of some of the encounters I enjoyed at DATE'09 in Nice.
**************************
Various voices of those who exhibited at DATE, or not ...
Walking across the Exhibition Hall in the Nice Acropolis Convention Center, I did not sense the crisis of exhibitors recently reference by an EDA pundit. It's true, there were visibly fewer exhibitors at DATE'09 than in past years, but the booths - all of the same size and texture - seemed thoroughly democratized, accessible, and conversational. Some vendors told me they had a "great show," while others told me "traffic is slow." Was this spectrum of comments any different than I've heard at conferences for many years now? Absolutely not.
* I met with Ghislain and Sylvain Kaiser in the Mykronos Café - CEO and CTO, respectively, of DOCEA Power. The company they founded in 2006, based in Moirans (France), provides tools for power management at the system level. The DOCEA strategy separates various concerns - functional, timing, power, and temperature - at early stages in design. The idea of optimizing to multiple metrics is not new to EDA. Applying the idea at the system level, however, is an emerging concept. These two young brothers bring years of experience to their company, Ghislain having worked at STMicro and Sylvian at Infineon. Their enthusiasm, joire de vivre, and product traction at big design houses will translate into a mandate: Watch DOCEO Power over the coming years.
* I also met with David Murray, CTO at Duolog Technologies. His terrific enthusiasm is also palpable. Based in Galway (Ireland), Duolog addresses needs at the system level with their Spinner, Bitwise, and Weaver tools, all knit together in the Doulog Socrates platform. Standing in the Duolog Booth, Murray said hardware issues that should be brought to closure in 2 weeks today, still take 12 weeks or longer. In Doulog parlance, these "elephants in the room" delay access to silicon and hence delay the tortuous (my word, not David's) process of software development. Although Doulog started out as an IP and services company, they are now a tool provider intending to solve this dilemma. The Doulog evolution mirrors the historic trajectory of many an EDA company; the Doulog tools, allowing customized granularity of views into the affects of high-level design decisions, mirror the future of system-level design.
* In the Press Room, I met with François Englebert, Vice President of Sales at CoFluent Design, based in Nantes (France). An EDA company of long standing, CoFluent is a significant contributor to the widespread move to ESL. CoFluent Studio is a "visual ESL architectural development solution" that provides performance analysis for hardware/software co-design. François is a serious-minded individual, proud of the 80 man-years of effort represented in CoFluent Studio. Early versions of the tool evolved, per François, to meet the expressed needs of Tier 1 customers, "leading to additional profiling and simulation capabilities. [Recently announced] Version 3 includes an improved Eclipse-based graphical interface. Either you start with a whiteboard, or Excel spreadsheet, and guess at the optimal architecture for your new product, or you start with our tool to guide you to the best implementation path." How does CoFluent position itself with respect to other ESL vendors? François advised that I stay tuned: "As we improve our ability to couple our tools with UML or other virtual platform vendors, you will see how we position ourselves with respect to other vendors."
* It's a perpetual pleasure to meet at DATE with IMEC, based in Leuven (Belgium). The IMEC message, viewed within the context of the highly collaborative European design automation community, is appealing - a combo of government funding, universities, and industry players to create a research center and attitude to drive innovation across a plethora of technical paradigms. This year, I met with Rudy Lauwereins, past DATE chair, and Katrien Marent, in the IMEC Booth. Rudy explained the process by which IMEC decides to move forward with a particular initiative. It's a 3-step process:
1) Ideation - The wild ideas appear, completely unstructured. We use a Web 2.0-type environment, so everyone can access the ideas and modify them, or indicate whether the ideas are long or short term, with or without commercial application.
2) Stage-gate funnel - We go through the business plan. Is there a unique selling plan, differentiation, competition, freedom to operate with regards to patents, a market, and a useful business model if we work with companies?
3) Execution - We check program progress at regular intervals. Can it continue? Should it be stopped? Should it be stopped, but in a way that allows our customers to keep working on the idea nonetheless?
Rudy said, "Tons of ideas come out of this process. Some are too weird. Some are too close to [current market offerings]. Some do not have sufficient differentiation. But some are quite good, and [inspire us] to move forward."
* Ian Mackintosh is President of OCP-IP, a standards body based in Beaverton (Oregon), but one with many European members. I met up with Ian in a Meeting Room, near the Mykronos Café. Ian was totally jazzed about the new SystemC TLM kit for OCP. With over 200 industry members, his non-profit organization is busy delivering "the only openly licensed, core-centric protocol comprehensively fulfilling integration requirements of heterogeneous multicore systems, and [facilitating] core reusability." IP reuse, models, and system-level languages - how quintessentially DATE!
* Back in the Mykronos Café, I met up with Douglas Pattulo and Jean-Christophe Longchampt, both of TSMC Europe and based in Amsterdam (The Netherlands). Although not exhibiting, TSMC threw a huge shadow at DATE with their April 21st Tech Symposium announcement, "a foundry-specific Integrated Sign-Off Flow, available now 65-nanometer designs. [The flow] tightly integrates all process-specific items including pre-qualified library and IP, selected EDA tools, production-quality flow, advanced design methodology, and TSMC foundry technology files that have been proven and refined over hundreds of applications." I emphatically told Douglas and Chris that their news only further convinced me that TSMC is morphing into either an EDA vendor or, more likely, the brains of a virtual IDM. They laughed, denied my allegations just as emphatically, and we parted friends.
* Also in the Mykronos Café, I met with Cadence Design Systems Europe Vice President Wolfgang Stronski, based in Feldkirchen (Germany). Neither was Cadence exhibiting at DATE - in good company with Synopsys and Mentor - but the three Big Players in EDA were well represented on panels and in Private Suites designed for companies wishing to meet with their constituents without funding a booth. Wolfgang and I spoke about a perpetual question in EDA: How does the EDA vendor reach into the customer's organization? At the senior management level? Mid-level management? The CAD organization? The individual tool user? Wolfgang said that answer is a complex one, and is answered on a case-by-case basis.
* Ranking high in the category of Very Cool Companies, I met with Kees Steenberger from GreenPeak, based in Utrecht (The Netherlands). Yes, GreekPeak were exhibiting at DATE, yes I met Kees in the Press Room, and yes the company's story is compelling. They're a fabless semiconductor, providing "mesh networks that are enabled by the use of energy harvesting devices," matching the electrical needs of the applications within which their products sit with ambient energy sources, particularly of the mechanical variety. Watch for more news in the coming years from GreenPeak as their products help drive the burgeoning micro-sensor market.
**************************
Observing the EDA ecosystems at DATE ...
The formal meetings described above left me thinking about the mindsets that have driven EDA in the past, and the mindsets that will drive the industry in the future. Although EDA has appeared somewhat dispirited of late, given the low valuations of the publicly traded EDA companies and a sharp decrease in advertising, ergo coverage, in EDA, the things I saw and the people I listened to at DATE completely re-energized my optimism for the industry. It's true the Exhibition Hall is evolving - not surprising, given the amount of information available online about everything and everybody - but the technology remains endlessly compelling and chock-full of unanswered questions and challenges.
Those impressions were reconfirmed in criss-crossing the Exhibition Hall and running in and out of the hallways of the Acropolis Convention Center. Case in point: Synopsys' Yatin Trivedi told me, when we crossed paths, that he was very enthused about the quality of the conference program at DATE, and the Exhibit Hall. He gestured over to the vibrant mass of people milling about in the University Booth and challenged anyone to say that a) DATE is over, or b) the EDA industry does not have a thrilling new generation of technologists ready to come of age.
I also caught up with the intensely over-committed Andrew Kahng, professor at U.C. San Diego, founder of Blaze DFM, and General Chair for DAC'09 in July in San Francisco. Ever the multi-tasker, Andrew was simultaneously carrying on a conversation with me, exchanging witty repartee with Synopsys' Anton Domic, and continuing to bang out emails on his laptop on the table in front of him in the Mykronos Cafe. Andrew and Anton together insisted that EDA is ferociously alive and full of technical lures that will pull in the next generation of technologists. Anton admitted, as (gently) accused by both Jan Rabaey (and me) during a panel earlier in the day, that the major players in EDA can seem complacent at times. Wrong conclusion, per Anton. The real problem is lack of in-depth coverage of the technology, the exciting innovations at EDA companies big and small. (Did I mention there's a shortage of ad dollars to support that coverage?)
Meanwhile, many others from the EDA ecosystem were also clearly visible at DATE'09: D&R's Gabrielle Saucier; Virage Logic's Yervant Zorian, doing yeomen's duty coordinating Wednesday's Special Day on SoC Development; EDAC's Bob Gardner, working overtime along with Anton Domic to coordinate Tuesday's full-day Executive Session; Mentor's Simon Bloch and Joe Sawicki; Tensilica's Grant Martin, busy winning the Best Follow-on Questions Award; Stanford's Subhasish Mitra; ARM's Rob Aitken; Synopsys' Tom Williams and U.C. Berkeley's Jan Rabaey, both winners of the EDAA Lifetime Achievement Award, Williams in 2007 and Rabaey in 2009; edaForum's Juergen Haase; multiple generations of DATE General Chairs, including EPFL's Nani De Michelli, University of Bologna's Luca Benini, Politecnico di Milano's Donatella Sciuto, and IMEC's Georges Gielen; DATE Press Chair Freddy Santamaria; Mentor's Anne Cirkel, EDAC Liaison; TIMA Lab's Bernard Courtois, EDAA Liaison; SAME's Anne-Claire Desneulin; and Mentor Graphics CEO Wally Rhines.
Why do I go to the trouble of providing this partial list of folks I observed or chatted with at DATE? Because again, if you think EDA is dead, or conferences no longer provide opportunities for the cross-fertilization of ideas within EDA, you are wrong. The need for face-to-face conversation has not diminished. People who work in technology need to feel the push and the pull of other people's ideas, live and in person. Conferences provide that opportunity in a structured, and surprisingly unstructured, way. Travel budgets notwithstanding, technologists still need to gather under one roof on a regular basis. It's the nature of the human condition, whether you like it or not, that people who need people are the techno-savviest people in the world.
**************************
Panels & sessions at DATE ...
My perpetual lament still holds true. Too much quality content running on parallel tracks makes it neigh-on impossible to see everything at a conference as large as DATE. I gave it my best shot, but best is never enough when it comes to DATE - or DAC, for that matter.
The opening session, MC'd by DATE'09 General Chair Luca Benini included numerous awards, acknowledgements, and celebrations of the quality and quantity of paper submissions for this year's conference; over 400 people provided paper reviews in assembling the program. Also on stage during the opening ceremonies: Linkoping University's Zebo Peng announcing the 2008 Best Paper Awards; University of Southampton's Bashir Al-Hashimi presenting conference program details; Bernard Courtois presenting DATE Fellows, and EDAA Award winner Jan Rabaey; Yervant Zorian presenting a slate of IEEE Fellows; and Wally Rhines accepting Aart de Geus' Phil Kaufman Award.
ARM CTO Mike Muller's keynote followed, a lively, creative romp through the last 25 years of microelectronics. Muller offered 4 truths to live by: 1) Formal proving, 2) Hardware guys cannot be trusted, 3) Systems dominate, 4) Every answer is 0.1 or many. Before he was done, we learned the world is becoming strange, Amdahl's Law is bogus, and software as a service, plus multicore, is great news for hardware designers, but not so great for the software guys. Anytime someone like Muller - rife with personality and a sense of humor - takes the stage, everybody should be in attendance. At DATE, it looks like they were.
A more serious and technically complex keynote was offered up by CNRS' Joseph Sifakis, a 2007 ACM Turing Award winner. Sifakis parsed his topic, Scientific Challenges in Embedded System Design, into Reactivity, Autonomy, Dependability, Functionality, Granularity, and Cost. By the end of the talk, the world knew it has its work cut out. Operating systems are too undependable and too complex. System architecture needs to be simplified, with improved observability. Thread-based programming must be the new paradigm, while constructability and predictability must become the norm. Sifakis's message was not an optimistic one. Too many problems, too few brains to solve them.
Oh well, off to the Exhibition Theater for some lighter fare. Nani De Micheli moderated "Consolidation, a Moor of Venice Tale," a discussion with Synopsys' Anton Domic, STMicro's Marco Montaiti, ARM's Mike Muller, and Mentor Graphics' Joe Sawicki, which cast the conflict between adopting new tools or sticking with tried-and-true EDA vendors as an Othello-like tragi/comedy of epic proportions.
ARM's Muller said the cost of infidelity is high. But ST's Montaiti said most EDA suppliers cannot offer a complete tool suite, while De Micheli said it could all be resolved with better standards. It fell to Domic and Sawicki to declare the EDA industry is not complacent. Per Sawicki, "We are still bringing young people into this market!" Per Sawicki, "We are sponsoring significant innovation!" So there!
Tuesday at DATE closed with another Exhibition Theater discussion, "Open Source Hardware IP," moderated by Sun Microsystems' Shrenik Mehta, currently Chair of Accellera. Mehta's company has just been purchased by Oracle, an interesting development for those who follow the Open Source community, MySQL in particular. That topic was not on the table during Mehta's panel, but skepticism was. The audience clearly was not fully on-board with the concepts and implications of Open Source. No matter that the professors on the panel, University of Arizona's Vazgen Melikyan, University of Texas' Jacob Abraham, and Europractice's John McLean attested to the usefulness of Sun's open source OpenSPARC architecture as a teaching tool.
Before leaving the Convention Center on Tuesday, I did pop in at the back of the room to hear OSCI President Mike Meredith addressing a packed house about developments and updates in SystemC. Although, I did not stay for the session, it was clear that the European design community is very committed to system-level design as a paradigm, and SystemC as the language of that paradigm.
Wednesday's focus at DATE for me was the late-morning panel in the Hermes Theater that I moderated debating the pros and cons of vertically integrated organizations versus a disaggregated industry of multiple players. Panelists included Wipro's Denis Audoly, TSMC's Chris Longchampt, Virage Logic's Yankin Tanurhan, and Mentor Graphics' Wally Rhines. Clearly, these people all have vested interests in a highly disaggregated industry, where IP providers, fabless houses, EDA vendors, and third-party foundries can provide products and services through partnering and/or competing with fellow companies in their various niches.
For the sake of argument, I chose to defend vertical integration, where everything from IP to tools and manufacturing data is readily available up and down the organization. Although I was playing devil's advocate with that stance, after reconsidering the discussion during the panel, I think there actually is lots to be said for the coordination and cooperation that exists within a vertically integrated whole. The fact that TSMC is beginning to wield so much power back up into the design flow adds to my conclusion that integration is a natural outcome.
Meanwhile, Rhines vigorously defended the historical swings, integration to disaggregation and back again, and said they were the natural outcome of disruptive technologies. Tanurhan said the IP industry has matured to the point that IP is no longer suspect, even if provided by a third-party vendor. Audoly said external providers of IP and services optimize investments for all involved. Longchampt said it's all about partnerships and resource management. I still say I'm not convinced. This disaggregation thing still looks like messy stuff, albeit it's the real world. A top-down approach only lasts for so long.
The next session for me was Engineering & Technology's Chris Edward's Exhibition Theater panel evaluating the move to 65 nanometers. Per Rainer Kaese, Toshiba's focus is 65 nanometers for digital RF designs, and 130 nanometers for analog/mixed-signal designs. They've skipped 90 nanometers. Trevor Robinson, of Desix Technology, said the incremental performance improvements between single nodes have not been motivating, hence skipping nodes has become the norm. EDA Tech Forum's Paul Dempsey posed a question from the audience, "To what extent does your progress through nodes reflect influence from your manufacturing partner?" SIDSA's Fernando Barberio said, yes there is influence. Trevor Robinson agreed: "We are intimately tied to choices suggested by foundries." Rainer Kaese added that Toshiba is only able to offer a variety of flows to customers at mature process nodes. Otherwise, the foundry partners do indeed influence the choice of flows.
[Editor's note: My coverage of Chris Edward's panel first appeared in real time on Twitter. Thanks to Chip Design's John Blyler for encouraging me to jump onto the Twitter Train.]
[Additional note: I was disappointed to have missed a session in the Exhibition Theater on Wednesday at DATE, moderated by Gary Smith. Initially, I thought Gary's panel was not in the printed conference material, hence my oversight. Later I saw the panel was in the program, but listed after the Friday workshops. Way too confusing for the average conference warrior!]
The next session in the Exhibition Theater was hosted by SAME, the Sophia-Antipolis MicroElectronics Association of southern France. Jacques-Oliver Piednoir, President of SAME, favorably compared the technical excellence and business opportunities for design houses choosing to pursue operations in Sophia-Antipolis with opportunities in India, China, or Silicon Valley. His argument: If you do not utilize quality design personnel, whatever savings you may realize in salary in less-expensive labor markets, you will lose in multiple re-spins of faulty designs. Per Piednoir, "First time silicon success is a key part of success for your company." Clearly SAME believes they make a compelling argument for companies to relocate to Sophia-Antipolis.
Closing out Wednesday, I attended a late-afternoon session on Health Care Electronics and heard two talks, Bart Volckaerts from the Cochlear Technology Center speaking about cochlear implants, and one on brain implants for deep brain stimulation as a treatment for Parkinson and other neural-type afflictions, this second work being done at IMEC.
Both systems have been developed with a recognition that the human body is not an ideal location for conventional electronics. Electronic devices want a dry environment; biological systems need fluids to succeed. Physiology and physics, ergonomics and social implications, not to mention governmental oversight of invasive technologies, all play a role in the R&D in these two research areas. Although this session was the last of the day on Wednesday, the room was packed, with people paying close attention to all of the details and asking many questions. Clearly, there is no complacency in the population engaged in pushing the envelope in health care electronics.
**************************
DATE with Destiny ...
By the time Thursday dawned at DATE'09, like many people I was beginning to run out of steam and/or attention span. Nonetheless, Thursday included a full day's treasure trove of discussion about Everything Multicore.
I was involved in industry conversations for a portion of Thursday (see above), but did manage to late-morning presentations from ARM's John Goodacre talking about multicore technologies, and NXP's Peter Kollig talking about heterogeneous multicore platforms for consumer multimedia applications. I have to say, despite all the doom and gloom from many speakers at DATE about how tough multicore is, or will be, both NXP and ARM talk as if multicore is already a reality in life. True, there are a lot of additional problems to resolve, but these guys are shipping products that are seriously multicore, or helping their customers to do it.
I actually started out Thursday attending the second half of Grant Martin's tutorial on multicore technology. Just as with the folks from ARM and NXP, Martin carefully articulated the real issues, and possible solutions, for those who are working to add new cores to legacy cores in increasingly multicore devices.
The lunchtime keynote at DATE was also in the spirit of the Special Day on Multicore. The CTO of STMicro, Eric Flamand, presented a detailed, almost academic talk on the complexities at hand. His argument: replace heterogeneous, mixed hardware and software solutions to a design problem with a multicore "fabric," the cores being activated as needed, the others left dormant. Flamand's implication, if I understood him, was that there is so much real estate on-chip today, why not lay it out in advance with multicore in mind, deciding after the fact how to utilize the capabilities there.
The afternoon sessions at DATE on Thursday included a panel discussion evaluating the role startups will play in driving multicore technologies. I only stayed for part of the session, long enough to hear Grant Martin say, to get funding "startups must offer significant differentiation over what is already out there. The last thing I want to hear is another me-too [offering], with only a 15 or 20 percent improvement over existing products."
As Thursday drew to a close, the Exhibition Hall also began to bring down the curtain on DATE. As booths came down, I went upstairs at the Convention Center to attend part of a technical session on mixed-signal and RF testing.
Tima Labs' H-G Stratigopoulos described a proposed strategy using machine learning to attack the problem of testing RF chips. Stratigopoulos said the Tima Labs technique required only 30 minutes to sample 1 million on-chip devices, and only 2 minutes to generate the required guardbands necessitated by the findings.
After his presentation, we spoke about the commercial applications of the technology. He said full implementation would require shutting down testing facilities to "re-tool" if this type of strategy is to be explored. I predict, given the promise of machine learning, that it won't be long before the strategy is fully utilized across the width and breadth of electronic device manufacturing and test. At that point, the topic will no longer be relegated to the last session on the last day at DATE.
Thursday ended on a high note for the Editorial Community attending DATE'09. Thanks to a gracious invite from the ARTEMISIA Association, we were treated to dinner and a concert on the top floor of the convention center. The 'concert' was actually a robotics contest, pitting student design teams from Australia, The Netherlands, and Sweden. The task was to build robots that could play instruments.
There were a variety of rules for the contest, a tough panel of judges, and big prizes - 3000 euros, 4000 euros, and 8000 euros for the third, second, and first place teams, respectively. The team from Eindhoven won, not only because their instruments were spot-on musically and could follow a human conductor, but because the electronics for one of the other teams failed during the performance.
The instruments in the various bands included a guitar, a drum set, a flute, an accordion, and a piano. The Eindhoven team had participated in the contest last year and was able to learn from last year's errors. The losers in this year's contest were new to the effort, but the prizes are so large, surely they'll be back to try again next year.
What a spectacular way to end DATE'09. Enthusiastic engineering students, robotics - mostly built on programmable devices, fascinating music, a kind of Robots Have Talent pre-dinner extravaganza, and a fabulous meal. It doesn't get much better than that!
Next year DATE'10 will be in Dresden. Much may change between now and March 2010, but no matter if you have to be pushed or pulled, you should plan to attend.