July 31, 2006
DAC 2006: As the world turns, the pendulum swings
Please note that contributed articles, blog entries, and comments posted on EDACafe.com are the views and opinion of the author and do not necessarily represent the views and opinions of the management and staff of Internet Business Systems and its subsidiary web-sites.
The world was in a muddle in the days leading up to the 2006 Design Automation Conference in San Francisco. The hostilities along the Israeli-Lebanese border intensified by the hour; the Israeli army appeared poised for an invasion. Factions across the Arab world seemed divided as to how to respond to developments; Washington did not. The temperatures across the length and breadth of California reached record levels; in particular, the mercury hit 116 degrees Fahrenheit just a few miles inland from way-too-sunny San Francisco. On Wall Street, Dell's stock tanked on news of severely reduced Q2 earnings. And, a grand jury investigating the baseball steroid scandal was extended, allegedly because there were some who still wanted to pursue an indictment of Giants' slugger Barry Bonds. Local, regional, national, and international news were all pretty agitating.
Strangely enough, the world was still in a muddle in the days after DAC 2006. You would think DAC might have shifted the tenor of the news from bad to good, but it did not. The horrific death count in Lebanon continued to rise - and in Israel, although at a slower pace. The Arab world seemed closer to consensus in their condemnation; the G8 did not. The deadly heat wave across California had subsided; the wind and fog had returned to The City, but over 120 deaths statewide were directly linked to the unprecedented temperatures. Dell's stock had rallied a bit, but the SEC appeared to be drawing closer to outright indictments of dozens of companies in Silicon Valley based on the backdated stock option scandal. Meanwhile, Tour de France champion, American Floyd Landis, faced the loss of his title should he fail the second of two mandatory drug tests, having failed the first for the presence of too much testosterone.
In the midst of the geo-political saber rattling and death, allegations of drug use, and increasingly alarming evidence of global warming, DAC was a blur - a cacophony of pre-dawn train rides, 20-hour days, a thousand handshakes, and few meals enjoyed to completion. I never saw anybody or anything from Cadence - they're were having an off-site, which I did not attend. I had a few EDAC hors douvres on Sunday night before the Dataquest slideshow; missed cocktails, but made dinner with Mentor Graphics at SF's MOMA on Monday where they celebrated the 25th anniversary of the company in style; briefly enjoyed Synopsys' hospitality over breakfast on Tuesday (before rushing off to the opening conference keynote) and drinks on Wednesday in the early evening. On Tuesday evening, I was a bit ill-mannered in walking out of the Magma/TSMC pre-dinner presentation at LuLu's because I had run out of attentiveness (in particular) and time (in general) for a presentation I had already seen; TSMC has integrated Magma's tools into its version 7.0 reference flow. My punishment - only a piece of bread and a glass of (sparkling) water for dinner.
I enjoyed Catalan cuisine with Fenix, Takumi, and ACE at Belden Place on Thursday, and had lunch courtesy of CoWare at the AMC Theater on Wednesday where I could only stay long enough to hear the first two customer presentations, so missed the punch line of the event; CoWare has released its SystemC Modeling Library source code and reuse methodology guidelines. I was there long enough, however, to hear a poor, exhausted engineer in the middle of the theater softly snoring over his half-eaten box lunch to the embarrassed amusement of those sitting nearby.
I made it to the Denali party Tuesday night and stood at the front of the dance floor in my business suit, holding my briefcase and marveling at the quality of musicianship from the likes of Gary Smith, Aart de Geus, Bob Gardner, Ted Vucurvich, and Jim Hogan. I managed to lob the harshest of criticisms at DeepChip at the Hacks & Flacks panel on Wednesday afternoon, only to find that Synopsys was celebrating Cooley's site at their elegant paté & champagne soiree in the Flood Mansion just 2 hours later. We all received a bumper sticker from Synopsys designed with Cooley in mind. It said, Synopsys - Chip Happens.
Meanwhile, ESL continued to press its case all week long, and although everybody at DAC seemed to have a DFM strategy, DFM continued to press its confusion. I managed to bungle appointments with DAFCA (first a no show, then late for the re-scheduled appointment), got appointments with Xyalis and OneSpin confused and was therefore late to both, got the location of the Critical Blue appointment wrong (missed it), arrived too early for ArchPro (missed that one, as well), and forgot Nano Solutions completely -- an astoundingly awful track record and my worst ever at DAC. I did, however, achieve a modicum of timeliness with CiraNova, Lynguant, TSMC, Tennison, BlueSpec, CEDA, SPIRIT, Synopysys, Forte, Sequence, and the SRC. Plus, I glanced in at the Tensilica booth to see the successes they'd achieved with their painting project, which raised money all week for the San Jose Ballet.
I moderated a panel on Tuesday in the DAC Pavilion with Cadence's Jan Willis, Penn State's Mary Jane Irwin, and IBM's Ellen Yoffa (winner of this year's Marie R. Pistilli Woman in EDA Achievement Award) that addressed the presence, or lack thereof, of Women in EDA; I moderated an Accelera panel on Wednesday, early in the morning at the Marriott, that gave five EDA user companies - ARM, Freescale, Intel, Sun, and TI - the opportunity to showcase their efforts-to-date evaluating and/or migrating to SystemVerilog as a verification language; I moderated a lunchtime panel on Thursday demonstrating an ESL flow from ARM to Mentor, facilitated by a SPIRIT wrapper-enabled IP handoff, but in the process completely missed a panel where the verification gods held court - Synopsys' Janick Bergeron, Cadence's Andy Piziali, and Mentor's Harry Foster.
I averaged 4 hours of sleep per night, passed on all alcohol on Sunday and Monday, had some on Tuesday, just a sip on Wednesday, and way too much on Thursday. Every other hour, night and day, was awash in caffeine; I sprinted at least 3 miles a day between conference venues, lost 5 pounds in the process, and ended the week by scrambling to assemble this copy.
But there's more: I was warned by Aart de Geus not to be naïve by publishing complaints that Cadence's earnings call was smack dab in the middle of DAC week; I avoided the Press Room as much as possible because the sight of editors sitting there rather than taking advantage of the host of sessions or panels offered every day at every hour seemed to rankle me more than usual; and with rare exception, I planned my week so badly that I was forced to arrive late and/or leave early to every single event on my dance card.
I allowed myself to be discouraged by the blatant lack of women technologists in the industry - with the notable exception of one technologist who is bravely enduring the transition to becoming a woman, although her contributions to the industry happened a long time ago, when she was still a man - and fought back on the battlefield of my own despair by engaging with as many academics as possible, particularly women academics, and concentrating on the talented women I met at the Workshop for Women in EDA. I actually dozed off during one of the panels I myself was moderating, managed to call a host of people by the wrong name over the course of the week, and repeatedly suggested to the Accellera breakfast audience that Yatin Trevedi worked for Denali, when I knew darn well he worked for Magma.
I found myself hating capitalism, loving academia, hating journalism, but loving journalists, hating snobs, but loving people with class, hating hubris, and loving humility, hating coffee, loving alcohol, and in the end - completely mystified as to how I could summarize the veritable jig-saw puzzle of impressions that comprised my week, and even further mystified as to why anyone would care what I thought or had to say about the whole jumble, jump, and jive.
Then, I came up with an idea. It's a cheap trick, but oh well. Why not post my Top Ten list of Top Ten lists? Embedded in there somewhere might be the truth about DAC. And what is that truth? You've probably got an answer, but here is mine: The truth is that DAC, like everything else, is about a pendulum that swings, from good to bad, from positive to negative, from here to there. To hell and back, the pendulum swings as the world slowly turns beneath it.
Back and forth, and back and forth. Peace prevails, and then there's war. Companies rise, and then they fall. Leaders climb to the top, and then are ousted, resign, or die. Technologies converge, only to diverge once again. Industries consolidate, only to disaggregate when everybody's forgotten why consolidation was critical. Publications dominate, then whither to a shadow of their former selves. Consumers clamor for that must-have device, then scoff & toss, and move their dollars to a different market sector. And in the midst of it all, abstractions rise, and sometimes fall, FPGAs duke it out with ASICs, self-assembling nanochips stare down CMOS, and pundits ponder whether the international language of business will still be English in a hundred years - or even 50.
You can find the full EDACafe event calendar here.
To read more news, click here.
-- Peggy Aycinena, EDACafe.com Contributing Editor.