Will Blog for Food
[ Back ]   [ More News ]   [ Home ]
Will Blog for Food

The way to a journalist's heart …

06/01 Feeding the Blog Monster in Cyberspace
05/31 Hors d'Oeuvres with EDAC in San Jose
05/16 Dinner with Mentor in San Jose
05/15 Cocktails with ARC & Toshiba in San Jose
05/15 Coffee with CAST in Santa Clara
05/08 Breakfast with Xilinx in East Palo Alto
04/19 Lunch with OneSpin Solutions in Redwood Shores
03/06 Breakfast at SNUG in San Jose

06/01   Blog Monster feeds on EDA Weekly in Cyberspace

I was copied on this e-mail yesterday. It was from ValleyPR's Georgia Marzalek and was addressed to ESNUG's John Cooley:

Subject: Fun Peggy stuff--Re: Cooley's Personal EDA Blog (May 2006 Digest)
Date: Wed, 31 May 2006 13:41:28 -0700

Hi John --

My trick for spelling that last name -- think Peggy Cinema, now add Ay to the beginning and change the m to n.

-- Georgia

Gosh, I thought, what the heck is Georgia talking about? So, I took a minute to open the link to John's industry-wide e-mail blast and - oh my!

Last week the infamous EDA Peggy (everyone calls her "EDA Peggy" because no one can remember how to spell her last name) published the results of her survey where she asked the EDA vendors: "What's going to be the 'buzz' at the upcoming 2006 DAC in San Francisco?".

She got 65 responses and, of course, 59 of them were the usual self-promoting EDA vendor infomercials pimping their own particular set of EDA tools. But, by some quirk, 6 of them slipped up and were accidentally honest!

Wow! In the category of “There's no such thing as bad press,” I guess I owe John for furthering the notoriety of both EDA Weekly and my odd last name (which is Basque, by the way).

But, to suggest that the entries in Buzz@DAC.2006 were not honest seems odder than my last name. Who's lying? Are the companies who sent in those 59 entries making that stuff up? They might be wrong - or even pompous - but that's not the same as lying. And, are those companies not really enthusiastic about the tools and the technologies that they're laboring over 50, 60, 70 hours a week, or more?

Also, to suggest that companies “pimp” their tools (ew!) seems even odder yet.

Look, companies have products, politicians have platforms, and journalists have self-promotion at the center of everything they do. That's not news. It's not illicit. It's about free markets, free speech, and freedom in general. It's not about pimping.

Pimping is about prostitution. It's about somebody owning somebody else's body and, by inference, their soul, and selling that body and soul to a john. But the third-party tool vendors don't own their products, body and soul, because the products are just products. They're not human beings. They don't have bodies. They don't have souls. And they don't suffer the way a prostitute suffers who's being held in bondage by a pimp.

The tools from the EDA vendors are just third-party CAD software applications that are used by digital, mixed-signal, and analog IC designers - and a system architect, here and there - to optimize some or all of a circuit design. There's nothing illicit or illegal about it. Designers aren't pulling up along side the curb in some seedy neighborhood somewhere to strike an unholy deal with a poor young tool who's being forced to stand out there by an evil pimp lurking in the shadows of the liquor store nearby.

CAD tools are nothing more than a few (million) lines of code written at some weird level of abstraction that reflect an algorithm (or several), which is an approximation of an intelligent, optimized way to solve a puzzle, which is eventually translated into a series of ones and zeros, which then tells some other ones and zeros what to do, which then gets translated up through yet another maze of abstraction to eventually produce a tiny little bit of semiconductor material, that's been assembled in such a way that, when subjected to an electric current coming in (usually from the left), produces a current (that exits to the right) that makes a device downstream from that bit of semiconductor material do something like ring, beep, click, turn on, turn off, toggle, twinkle, speak, or honk.

It's kind of cool, actually - a lot more cool than pimping or prostitution. Making, or using, CAD tools to design stuff is something you can be proud of. It's something you can write home about. Or blog about …

05/31   Hors d'Oeuvres with EDAC in San Jose

The EDA community - or at least the EDAC sub-set of that community - gathered tonight at the Techmart in Santa Clara to hold elections, and to hear about the endless, annoying, overwhelming governmental, legal, accounting, and export regulations that are crowding everybody's style, consuming funds that could otherwise be used for innovation and R&D, and taking some (but not all!) of the fun out of the wild and wacky world of high-tech start-ups.

But first - to soften the pain - wine, some hard liquor, a hot buffet (the pot stickers were my favorite) and plenty of schmoozing were all on the menu prior to the presentations. And after cocktails, there were the 2-minute speeches from the 13 candidates running for the 9 spots on the EDAC Board of Directors. I think those presentations helped to remind folks, in advance of the buzz-killer legal and accounting presentations, that EDA continues to be a great place to hang out.

People in this industry are just darn interesting - particularly when they're running for Student Body President. I don't want to add salt to the wound of those who ran for office tonight but learned they lost after the ballots had been cast and counted, so I'm just going to paraphrase some of my favorite parts of the some of the speeches - without attribution.

“The EDA industry is a place that makes a difference in how electronics moves the world.”

“There's lots of work to be done.”

“EDAC is an actionable organization. When we decide to do something, we can get it done.”

“Consortia like EDAC need to help cross-pollinate with other industries, so we can share ideas that work.”

“We need more investment to promote more R&D and innovation in EDA.”

“We need fresh blood in EDA. We need to welcome the next generation of entrepreneurs.”

“Innovation happens at the edge of an industry. We need to advocate for small companies.”

“The small companies in EDA can't do it by themselves, but working within EDAC we can get it done.”

“We can't deliver value as an industry unless we cooperate with each other.”

“EDAC is the watering hole of the industry. “

Looking around the room at the folks holding their drinks and plates of appetizers, I think that last was the most compelling statement of the evening.

(By the way - EDAC Treasurer and CFO, Bob Gardner, assured the Press that the list of new EDAC Board members would be on the EDAC website by morning.)

05/16   Dinner with Mentor in San Jose

Boy, that was fun! Mentor Graphics treated us to a fabulous meal at the Hilton Hotel in downtown San Jose, and then walked us over to the historic California Theater - what a beautiful building! - to hear Dr. Bill Saturno from the University of New Hampshire talk about the origins of the Maya.

It was really neat to meet him during cocktails, and then to hear his lecture. Dinner included steak and lobster, plus some wonderful wine. I continue to think where there's food, the Press aren't far behind.

Dr. Saturno's research at San Bartolo in Guatemala over the last decade was profiled in National Geographic in January. He says that the marvelous and sophisticated mural he unearthed in a complex of buildings at San Bartolo - it depicts the Mayan creation myth - dates from a far earlier period in the history of the Maya than experts would previously have believed possible.

Saturno's discovery and subsequent work at San Bartolo has made many experts re-consider when the Classic Period of the Maya actually took place. Instead of falling between 300 CE and 900 CE, perhaps the Classic Period of the civilization was already underway by 100 BCE - the date of the San Bartolo mural.

Saturno also talked at length about his work with commercial satellite imaging to map and predict the locations of other Mayan sites in Central America and Mexico. He said the Maya used a lot of lime-based 'paving' to create their cities and those materials have had a lasting effect on the flora of the area in the vicinity of the population centers - even today, 1500 to 2000 years later. Those areas can be distinguished from their surroundings, even if the structures themselves are buried, because the coloration of the flora is different as seen from space.

During the Q&A after his talk, Saturno agreed that the effect on climate and life forms that we suffer from in the modern era due to the 'paving' over in huge metropolitan areas could have been predicted had we better understood what appear to be the negative affects of the Mayan metropolitan areas on their environments and populations.

05/15   Cocktails with ARC & Toshiba in San Jose

The ARC/Toshiba Press & Analysts event was heavily attended today. Standing at the back of the room and counting, I figure there must have been at least 75 people there, plus a camera crew, a stage upfront complete with bar stools and executives from ARC International, Toshiba, and Cadence, and big screen TVs for those of us in the back who were straining to see.

I arrived late because of my coffee with CAST, so things were fully underway when I reached the Silicon Valley Capital Club on the 17th floor of the Knight Ridder building in downtown San Jose. There's no posher venue in all of Silicon Valley, so not surprisingly, the food laid out in the back of the room looked exquisite.

People standing in the back with me - other folks who had also come in late - were busy looking through the materials we'd been handed, just as I was. The CEO of ARC was up in front talking. Head honchos from Toshiba were there. A Senior VP from Cadence was there. They were speaking, answering questions, and looking in command of things.

In back, we figured out that ARC and Toshiba assembled this meeting in order to announce a “strategic collaboration intended to grow the worldwide semiconductor industry's adoption of configurable technology.” Toshiba has signed up for a multi-year license to use ARC's ARChitect processor configurator. The companies say they're also going to work together on a new version of ARChitect that will be custom suited to Toshiba's MeP Media embedded Processor. Cadence was on hand to endorse the news and to say this is a very, very good idea.

During Q&A, a question from long-time journalist Nick Flaherty went something like this: Why does Cadence need to be here? Isn't the key thing here the MeP ARChitect announcement? Will there be room in all of this for Magma, Synopsys, or Mentor Graphics?

The answer from ARC went something like this: We've aligned ourselves with Cadence because we value people who see problems the way we do. The technology is only as good as the problem you solve, so for us it's important to partner with people who really understand the problem. The meat of the thing is to flesh out the problem. And leaders, by their nature, aren't afraid of competition.

Hmm, I thought - that sounds great. A guy standing next to me whispered, “I think this is just a licensing announcement. This is a heck of a gathering for ARC just to announce a customer.” “Who's complaining,” I asked him. “Look at this food. Look at that view!”

After the Q&A, taping, and general session concluded, everybody in the room was invited to adjourn to the balcony of the Club to enjoy food and beverage and ambiance. The balcony overlooks all of the Silicon Valley. It was spectacular out there this afternoon, plus they had a jazz ensemble and waiters passing around wine and great appetizers. What a life we in the Press have, I thought. There were a lot of other journalists out there partaking and imbibing who looked like they agreed with me.

05/15   Coffee with CAST in Santa Clara

You gotta give it to Hal Barbour, president at CAST, for enthusiasm and joie de vivre. Have you ever seen a guy so excited about life, technology, or IP? I only had 45 minutes for coffee with Hal at the Double Tree Hotel this afternoon, along with the executives from Cortus, SA, who together are announcing a new line of 32-bit microprocessor cores they're calling the APS family of cores, which stands for Advanced Processing Solution.

CAST and Cortus are making their announcement this week because they're in town - Hal from the East Coast and the folks from Cortus in from France - to attend the In-Stat Spring Processor Forum. Because I was in a hurry this afternoon, and therefore can't read my own handwriting from our meeting, I've decided to take this canned quote from Hal right out of the Press release I was handed at the hotel:

“We've helped hundreds of designers succeed with 8051s and other controllers, and that market's clearly continuing to grow. But some systems just need more horsepower. Those designers have had little choice but to live with the technical and business overheads of an advanced 32-bit processor without actually using all of its capabilities. Now these APS cores provide an excellent 8/16-bit upgrade solution, and we're excited to be bringing them to existing and new CAST customers.”

You know, I don't really care that this is a canned quote. I can actually see Hal Barbour saying this - in fact, if my scribbled notes could be deciphered, I think he did say it. I also think that rather than sell IP, CAST ought to try to bottle and sell Hal's enthusiasm. It's contagious.

Unfortunately, although I was offered a meal by my hosts, there was no time. Paul Lindemann of Montage Marketing, who orchestrated the meeting today, did however slip me a chocolate chip cookie as I was hurrying away. I gobbled it down en route to my next stop of the afternoon.

05/08   Breakfast with Xilinx in East Palo Alto

Have you been to the new Four Seasons Hotel on 101 at University Avenue in Palo Alto? That's one fancy place - understated, sleek, quiet, and a great venue for Xilinx to hold a “Media Summit” and announce its new 65-nanometer Virtex-5 family. The breakfast buffet was elegantly presented, and it was a hoot to check out the conference facilities and watch the carefully choreographed event unfold.

Unfortunately, I could only stay until 10:30, so the parts of the summit that I caught were only the 30-minute welcome from Xilinx CEO Wim Roelandts and the 60-minute technical presentation detailing the new product and the new ExpressFabric technology from Xilinx.

Everybody from Xilinx was very dressed up, as were folks from partner companies standing along the back wall - including Synplicity, Mentor, Magma, and VMETRO - who were there to present, as well.

By the way, have you ever noticed that editors and journalists seem to hail from a different fashion planet than high-tech execs? The Press frequently look like they've slept in their clothes, or got dressed in the dark because the tie (if there is one) often doesn't match the shirt or the jacket or the slacks. Be that as it may, high-tech execs seem to be able to look past this small eccentricity on the part of the Press and be courteous to journalists nonetheless.

So, back to the Xilinx event: Xilinx says that customers are asking for higher performance, lower power, lower system cost, and a shorter design cycle. Guess what - Virtex-5 LX FPGAs are set to satisfy all of those requests. They're doing it through innovations in the process, the architecture (that's the ExpressFabric part), the hard IP (lots of stuff there), the packaging, and new design tools that play nice with tools from Synplicity, Mentor, and Magma.

Now, I'm not writing these things down - extracted as they are from the materials I brought home with me - because of the free breakfast, although my being interested might have been influenced by the food. I'm writing them down, because I want to point out that the most interesting parts of the morning were not the details about the Virtex-5.

The most interesting part was a comment made by Wim Roelandts before the technical tutorial. Wim noted that everybody now knows that structured ASICs have not lived up to their promise, and therefore we're back to understanding that FPGAs will be the only thing in the future when it comes to spunky, flexible, non-ASIC design paradigms.

I wanted to ask Wim, where have I been to have missed this fact? I know LSI's RapidChip structured ASIC initiative has been shelved, but I didn't know we're giving up completely on structured ASICs. So, I was glad I went to the Xilinx breakfast because that kind of info is even more stimulating than a strong cup of coffee served up with muffins and fruit.

04/19   Lunch with OneSpin Solutions in Redwood Shores

The Mistral is a stone's throw from Oracle's Emerald City in Redwood Shores, so it's a restaurant with a trendy and upscale atmosphere. What better place to meet up, then, with the Europeans from OneSpin Solutions who are in town touting their upcoming launch in early May? I had the carrot soup and a glass of wine, but neither OneSpin CTO Wolfram Büttner, nor VP of Sales & Marketing Thierry Le Squeren, were drinking - it was before 5 pm, for goodness sakes.

Besides, Thierry arrived late - by taxi straight from the airport. His plane had been delayed coming in from Paris, plus the mayor of Paris was onboard his plane and had to be greeted at SFO by the mayor of San Francisco. That may have helped with World Peace, but it didn't help Thierry one bit.

Anyway, Wolfram, Thierry, and I had a far-ranging conversation about the company, how they're a spin out of technology from Infineon and Siemens, and what they think the future of verification really is. Clearly, these guys think they're going to be a part of that future, and chalk it up to the wine if you want to, but their story sounds convincing.

The OneSpin product is an “automated static formal verification” tool (that's from the Press Release) and it's been used successfully, they tell me, by various European design teams. Now, it's time to launch the company and prepare to enter the EDA industry as a full-blown player - expanding geographically from their headquarters in Munich to having a presence in Silicon Valley, as well.

I'm thinking it would be good to track down the folks at Jasper, plus Harry Foster who's now at Mentor, to have further conversation about this formal verification thing, and to get their read on the OneSpin announcement.

Meanwhile, it was reluctantly revealed to me this afternoon that the OneSpin offices in Munich are very close to the fair grounds for that city's annual legendary Oktober Fest. Clearly, if the guys from OneSpin can be that clever in situating their offices in Munich, they can't help but be a success in EDA!

03/06   Breakfast at SNUG in San Jose

Synopsys has not always invited the Press to attend their User Group meetings. In fact, it's a habit they're just now getting around to. So, I was happy to make my way to the Marriott Hotel in Santa Clara to attend my first SNUG today.

The nice folks at the registration desk in the hallway outside the main ballroom were busy - lots of attendees, lots of badges, lots of materials to be distributed - but they were very welcoming when I arrived and had my badge waiting for me, plus a special request to sit in the “Reserved for Press” section in the ballroom nearby where Synopsys CEO Aart de Geus was going to be delivering the opening keynote for the 3-day conference.

Much to my liking, the Press section was not in the front row, but off to the right and mid-way back in the ballroom. I really like the anonymity of being back in the crowd when I'm attending industry events. Sitting in the front seems counterproductive to watching the way a speaker interacts with an audience.

Anyway, as I waited for Aart to begin his talk - he was introduced by SNUG San Jose Technical Chair Leah Clark from Broadcom - I noticed that my packet of information said: “Press are respectfully asked not to ask questions at the keynote. We will make Aart available for 15 minutes in the Press Room to answer questions after his talk.”

Aart de Geus spoke from 8:45 to 10:00. It seemed like he touched on every single thing that Synopsys is currently involved in. There were maybe 700 people in the audience. I didn't actually count them, but there were a lot and they all seemed to be paying close attention to Aart.

Some of Aart's slides seemed to suffer a little bit from what I call “happy talk” … but it's probably Aart's prerogative to put an optimistic spin on what the company's doing, especially at his company's own users conference. Besides, it looked like the hundreds of Synopsys tool users sitting around me were smart enough to know that tools from any vendor, including Synopsys, sometimes don't live up to the marketing pitch that precedes them.

After Aart was done, the floor was opened up to questions. The first question asked if mask costs going up are going to drive the number of design starts down. Aart said no, the scare is overblown. Masks are expensive, but the cost of design is coming down. Then, there were a couple of questions about test that Synopsys Senior VP and GM Antun Domic came up and answered. Then John Cooley posed a question from the floor.

He wanted to know why users seem less enthusiastic about Synopsys' IC Compiler than Design Compiler, and suggested that Synopsys had released IC Compiler too soon. Antun said no, the release wasn't premature and cited STMicro as one of several high-profile customers who are pleased with progress on the product.

Then Sunburst Design's Cliff Cummings asked a question about the ESL market. Aart said ESL promises to be a great market, but Synopsys' focus is on building IP because the ESL market is still very small.

John Cooley asked another question. Is Synopsys' Pilot Design Environment - John called it, “Your Pilot thingy” - just a hook for the company's design services or is it a real product?

Aart said, “We don't sell thingy's.” The audience laughed, but Aart didn't.

After the session was over, I hurried to the Press Room several corridors away in the hotel. I was hungry and I knew there would be good food there. I wasn't disappointed - a buffet fit for kings.

Aart eventually showed up and gave the Press Corps some minutes to ask questions, then rushed off to other meetings. I asked Aart about DFM and he said designers understand DFM but want the tools automated so they don't have to think about DFM. But, I really wanted to ask Aart why John Cooley was allowed to ask questions after Aart's keynote.

Isn't John part of the Press Corps, and wasn't the Press supposed to zip it until afterwards? But maybe I'm wrong. After all, John wasn't in the Press Room scarfing down food with the rest of us - when it comes to SNUG, John's a user. I guess it's only at DAC that he's Press. Also, maybe if John doesn't accept as much food from the industry as the rest of us, he's probably at greater liberty to be candid.

For more discussions, follow this link …