Profile: John Cooley of ESNUG
Cooley discusses Cooley
|
|
By Peggy Aycinena, Editor of EDA Weeky, June 2, 2003
|
|
|
John Cooley
|
Lots of folks in EDA like to throw around first names: Ray, Aart, Wally, Gary, Bernie, Penny, Jacques,
Rajeev - and of course, Joe.
But there's one icon in this industry who's definitely a last-name kind of guy, and that's Cooley. It's been a long time in coming, but I've finally had a chance to profile John Cooley and I have to say that, one-on-one, he's not half as scary as people would have you think. And though his raw candor during our conversation was not totally unexpected, his self-effacing charm caught me completely off guard. So please drop all of your pre-conceived notions and read on as Cooley discusses Cooley. -- Peggy
"I grew up on Air Force bases. My dad was a fighter pilot and served in Vietnam. I also remember as a kid at one point in the 1970's, we were living in Iceland. Weird place. Iceland is so far north that it has almost 6 months of total darkness in the winter and 6 months of almost total day light in the summer. It would be 11:00 PM at night and the sun was still up. Weird. Dad flew F-102 fighters against Tu-95 Russian Bears. The Bears were Soviet bombers flying from Russia down to Cuba and back. Dad was escorting them. My father retired from the Air Force when I was half way through high school and we moved back to his small town in Vermont. His small town became my small town and I graduated from high school there."
"I had a lot of fun times in high school, played football, did track, was captain of the math team - some early nerdiness there. They taught COBOL in my high school, but the school couldn't afford a computer, so you'd write a program and the football coach would look it over and decide whether or not it would run. That was in the late 1980s."
"From high school, I went on to the University of Vermont. I was there for about a year and a half, mostly partying my brains out. The ratio of men to women in that university town was 1 to 2 -- a dangerous ratio for a horny young Cooley. Consequently, I flunked out and joined the Army. At that time, you could enlist for just 2 years."
"I ended up working in a nuclear weapons unit while I was in the Army. It was very interesting, and even entertaining in some ways. The Army trains you to locate any place in the world down to a resolution of 10 meters - that's an 8 digit coordinate. We joked about needing only 6 digits - something like within the nearest half-mile - because nukes can't really be accurate down to the nearest 10 meters."
"After I finished my two years in the military, I went back to college and got straight A's from there on in. I wasn't a trust fund baby; the Army paid for my college education."
"One of the fun things I did during my college summers was to intern at the Argonne National Laboratory outside Chicago. Since I had been stationed just outside of Chicago while I was in the Army, I knew the area quite well. It was a lot of fun. I spent my college summers at the Labs designing electronics for the high-energy physics research going on there. They were trying to detect a neutrino decaying because it would verify some kind of Grand Unified Theory some physicist had. Normally at Argonne Labs they would only have hired physicists for everything, but for the electronics work they broke down and hired electrical engineers like myself. Physicists don't know electronics like EEs do. Since I had my nuclear weapons background from the Army, when I was doing high-energy physics stuff with the people at the Labs, they just loved it. Actually, I ended up doing DSP work there way before most people were doing it anywhere."
"I eventually graduated from the University of Vermont with a degree in electrical engineering, but it was a no-name school. It was tough to get a job as a chip designer coming out of the University of Nowhere, so I hired three freshman girls to address and stuff envelopes with my resume - at a nickel per envelope. I sent out 1,105 resumes. Do the math if you want to know how much that cost me. I got tons of rejection letters and many interviews. Some were just practice for jobs I really didn't want, but I did end up getting two jobs offers that I really did want."
"One offer was from Hewlett-Packard in California. The other was from Data General in this place called Massachusetts. The problem was my girlfriend. She was still in school in Vermont and her family was in Massachusetts. Of course, we ended up breaking up 3 years later, but because of her, I choose Data General at the time."
"Data General had a shoe-string budget in those days, so the engineers had to be everything there. It wasn't like other companies with departments. At Data General, we were all hands on in the lab. There wasn't a big staff of technicians to help."
"We were designing in ECL - the big, high-speed stuff at the time. That's when I had my first exposure to this language called Verilog, and to this weird thing called Design Compiler from this little 'nothing' company called Synopsys. And all of this was involved with this little weird technology called ASICs that no one thought would go anywhere. We were designing computers, and an ASIC was just part of the process."
"But then something went really wrong at Data General. There were all sorts of problems and financial turmoil. I'd been at DG for two years when I got chopped with a bunch of other people. But I quickly got picked up by Sequoia Systems which made fault-tolerant computers used for banking systems. Banking computers could never, never make a mistake and had to stay up and running at all times. If mistakes were made, the system had to self correct. That's what fault-tolerant was. Anyway, I used even more Verilog and more DC in designing gallium arsenide chips at Sequoia. Vitesse was our fab."
"It would have been a nightmare for our projects if Design Compiler messed up and we were finding its errors all the time. We began to be afraid to tape-out our chips, because a mistake meant a schedule hit and those nasty NREs."
"So I flew out to California with my boss and my vice president of engineering to visit the 'Mother Ship' for Synopsys. We met Harvey Jones, who was CEO at the time, and others. They all just sort of shook our hands and laughed. We asked for a complete list of bugs and they said, 'You're crazy. We don't share that information with anybody.' My boss and my VP of engineering were very frustrated because it left us flying blind. We had no idea what strange things DC was going to do and we had bet our company on it for our designs. The three of us went back home to Boston very frustrated."
"A couple weeks later, I went to a local Synopsys training class here in Boston. It was there that I discovered that I got more out of talking to other Synopsys users, sharing their bugs and workarounds, than I did by sitting in class. I took a piece of paper and passed it around the room, asking for names and contact information. The piece of paper ended up with 11 names and e-mail addresses on it, which became the East Coast Synopsys Users Group - the original meaning of the ESNUG acronym."
"The first 40 e-mails that we exchanged were just attempts to find a convenient place to meet. But it was too hard to do, so I finally just said, 'Screw it. Let's just do tech talk by e-mail.' Even today, if you look at the ESNUG archives, you'll see that ESNUG post 40 is the first one where I actually write about a bug in DC, and provide a workaround for it. Today we're at ESNUG post 412 and 17,000 subscribers."
"Anyway, news of ESNUG spread around by word of mouth. After the initial 11 members, the list picked up another 25 people. We went from 11 to 36, but we still weren't getting much discussion going. So, then I put the word out on the hardware design newsgroups: 'Hey. I'm doing this mailing list of user workarounds and bugs for Design Compiler.' I had 360 people sign up in just one day. That was in 1991. Back then, 360 e-mails was definitely a shocking amount of e-mail to receive in one day. The e-mail administrator at Sequoia was totally overwhelmed. He came to my cube and said, 'You're really taxing our system here.' But he adapted and we continued."
"That first day of 360 new ESNUG members also garnered 6 personal phone calls to me from Synopsys. First I had a call from the local AE [applications engineer] for the company, then the regional AE. Then I had a call from the local sales guy for Synopsys, and then the regional sales guy. Then the North American AE manager. Finally, I had a call from their VP of customer support. They all had the same message: 'You've got to shut this thing down.' Clearly, Synopsys was shocked that customers would be openly discussing bugs in their software."
"I responded to all of them the same way. I said, 'Look at the beauty of this thing. If there's a problem this week, by next week we all get the work around because of this dialog. I've got 400 people on the e-mail list; this saves you 400 support phone calls.'"
"But Synopsys didn't like it. Their legal department sent a letter to the CEO of Sequoia saying I was violating the NDA [non-disclosure agreement] with Synopsys. Of course, Synopsys didn't tell me they had done this. Just suddenly one day, my CEO walked into my cubicle and said, 'John, I don't know what you're doing, but you're going to talk to our legal department right now.' Then he handed me the Synopsys legal letter. My heart was going ba-bump, ba-bump, ba-bump. I was like: 'Oh my god! Why is the CEO of my company stepping into my office?'"
"I read the [non-disclosure] agreement, the parts that Synopsys legal had highlighted. But the more I looked, the madder I got. I wasn't reverse compiling the code, I wasn't violating anything. So I went to our legal department and explained everything to them. They, in turn, went to Synopsys and said, 'He's not violating anything here.' Things got more complicated after that."
"It was weird. Some executives at Synopsys would talk to me, while others would not. They would be friendly at one point, and then lawyers would be snapping at me the next. We started to have weekly conference calls within a few weeks of my forming ESNUG, calls between Sequoia in Massachusetts and Synopsys in California. At the time, it felt to me like Synopsys was trying to crack down on ESNUG and nothing more."
"My company's lawyers said that in this country there's still this thing called the First Amendment. The user's group wasn't reverse compiling. We weren't trying to steal any software. We were only sharing tips on using and debugging Synopsys tool. (I'm thankful I started ESNUG at a small start-up. If I was at a bigger, more political company I surely would have been crushed by my own company's internal politics.)"
"So I called this guy named Richard Goering at EETimes. And I said, 'I'm forming this users's group, but Synopsys hates me for it.' Goering was interested in a good story and this one was very colorful. The very next week, a teeny 3-inch story appeared in EETimes. The moment that story appeared, those executives at Synopsys that had been really nasty were suddenly very, very friendly, and Synopsys backed off."
"So our discussions with Synopsys continued. I'd be talking to their VP of customer support and I'd say, 'I'd loved to have your R&D people in on these discussions. They'd say, 'Yeah, yeah, yeah,' and do nothing."
"Now it's 1992. I thought it out and decided that the people upstairs at Synopsys were purposefully locking out their R&D guys from me, and I didn't know how to get to them. So I made a list of 110 names - all shots in the dark. There was bob@sysnopsy.com, john@synopsys.com, rick@synopsys.com. I made the list just guessing as what names might be there and sent each of them an invitation to join the ESNUG discussion. This worked like a charm. Most bounced, but I had reached some of the R&D guys and gotten past Synopsys management. The R&D guys wrote back and said, 'This is great!' Now I could talk to god himself - the guy who wrote Design Compiler. Before it was like getting an appointment with the Pope - lots of run around and permission asking. Now I had the direct line itself! Whoa. And these Synopsys R&D guys were really welcoming. Of course, Synopsys management was furious."
"I'd say dealing with Synopsys was like dating a schizophrenic woman. One part loved me and totally understood me. I was just a customer trying to use the tool saying, 'Please help me use the tool.' They respected that - engineer to engineer. I wasn't trying to drag Synopsys down. I was just trying to get straight technical conversation."
"But other people in Synopsys were just furious. Their salesmen were saying this is really a bad thing. Their legal department was going nuts, upper management was furious. They were sure that if you were discussing a bug, it would hurt their business."
"But I went, 'No, it's just realistic software usage.' You've got to understand that at that time, none of the EDA vendors discussed their bugs. They wouldn't say a word. You'd be flying blind using their software and you'd think that you were the only one who ever ran into a specific bug. They'd say, 'What's wrong with you? You're the only one who is seeing this bug.' They isolated us customers with classic divide and conquer tactics."
"Meanwhile, at my company, the CEO of Sequoia had been booking sales that did not exist. This was back in 1992/1993. For some reason, the SEC really didn't like that. Suddenly in a two-week period, The CFO suddenly left to 'pursue other interests.' Then the CEO left to 'pursue other interests.' The company had employed 450 people up to that point, but in a two-week period it went down to 120 people. There was wave after wave of layoffs. My manager laid me off and a bunch of other low-level guys at 9:00 in the morning. At 10:30 in the morning, he got laid off by his boss. Then his boss was laid off by his boss. We were all going to the layoff center in waves."
"It was devastating. And that's what made me an independent consultant. I wasn't losing my job because of technical incompetence. I was losing my job because some guy upstairs who made 20x my salary had screwed up and I was paying the price. But you know what? That first day being laid off, I had phone calls from three companies. They all knew me from ESNUG. I've been an independent consultant ever since then. I'll never go back to being somebody's employee. I don't like the insecurity."
"Of course the day I got laid off I had one of my regular weekly Synopsys conference calls scheduled. I left them a voice-mail to call me at home. Synopsys left me a cheery voice-mail back that said, 'We assumed our meeting was cancelled. You're no longer a legitimate Synopsys customer, John, and therefore ESNUG is dead.'"
"The funny thing is, I already had other e-mail addresses set up, addresses that were independent of Sequoia. I had nerd friends who had their own e-mail servers and they said, 'Here - we'll give you an account.' At that time only big companies and nerds had e-mail servers; the Internet hadn't gone AOL yet. I set up multiple accounts with different friends. That way, if Synopsys knocked me one out of one account, I had another one to fall back on. It was straight out of a guerilla warfare manual I had read in the Army. ESNUG ran on the communist cell model. The Synopsys people were furious, but I never stopped publishing. Years later I had heard the internal Synopsys management reaction at the time was 'Cooley is like cockroaches! You can't get rid of him!'"
"So I'm running this guerilla organization, but couldn't get my hands on Synopsys manuals. Remember, officially I wasn't a Synopsys customer any more, so users would secretly mail me Synopsys manuals to my post office box and say, 'You didn't get this from me.' Of course, I would always send them thank you's for this guerilla support."
"Then one day everything changed. Aart de Geus became CEO of Synopsys. Aart's new customer service guy, Bob Dahlberg, moved in and said, 'This is just crazy!' Bob had a philosophy that was 180 degrees away from the old school Synopsys paranoia. He said, 'We should be working with customers. ESNUG is a good thing!' Bob helped to change the perception that ESNUG was an anti-Synopsys movement, and convinced his management that it was a pro-Synopsys-user thing. He created the Ace Program to give customers the same training that the Synopsys AEs got, and established SolvNET."
"Bob believed in the open kimono, so that users could really see the tools, and not just in marketing terms. This was a very important change at Synopsys. I think, if they hadn't changed, the ESNUG paranoia would have stalled Synopsys. It's really clear on ESNUG, that if customers like a specific EDA tool, ESNUG helps it thrive. If they don't like something, ESNUG very quickly either kills off the tool or gets it back on course. Real user feedback always works much better than believing your own B.S."
"One of the neat things about running your own guerilla organization is
that you write your own rules. There's no boss to frown at you. I live on
a sheep farm that's 45 minutes west of Boston. It's a big old farm house
built in 1752 as the town's poor farm. Before there was Social Security
or Welfare, if you were down and out and you had no family to take care
of you, in frugal New England they used to send you to the local poor farm
to work for your room and board. This is why I have the Holliston
Poor Farm as my mailing address. It's not a joke; it actually was the
town's poor farm."
"Anyway, at any given momement we'd have 15 to 30 sheep at the farm. One
day the landlord decided to get two baby goats for variety. (He was a
newbie play farmer who didn't know how much trouble goats can be.) Within
two weeks my landlord was getting phone calls from angry neighbors that the
new goats had jumped the electric fence and were eating their garbage. My
landlord complained that "these goats have been trouble for me since
day one!" It was at that instance I asked my landlord if I could name the
two goats Aart and Harvey. He looked at me perplexed, not understanding,
but said "Odd names, but sure, if you want to. Why not?" The funny thing
is that I don't think my landlord ever knew that he had two troublesome
goats on his farm named after the CEO, Aart de Geus, and Chairman of the
Board of Synopsys, Harvey Jones. He just associated the names Aart & Harvey
as being a sporadic source of nuisance in his life -- just like I had."
"So [for many years now], I've been working in a purely consulting venue. If a customer has a project that's in crisis, they call me in. And because I'm the ESNUG guy, I get incredible support. People know me and fly me in. In fact, that's how I met Cliff Cummings. I was helping Tektronix out of a crisis and Cliff was a trainer at Tek then."
"I remember, many years ago, a customer phoned me up on a Thursday. They wanted me in California by Friday - they were in that much trouble. They had to get their chip out the door, and had to be ready for some big conference which preceded their IPO. The problem was that their synthesis runs were taking a week."
"I negotiated to not have to get there until Saturday. I arrived in Anaheim, and started working Saturday night. Their CEO met with me, along with their head of CAD support, and their engineers. They all had that really stressed look in their eyes - they were going nuts. I saw some of the stuff they were doing, including using some IP from third-party vendors that they didn't know how to incorporate into their chip. I showed them some tricks to get the weeklong synthesis runs down to daylong runs. I ended up working with them for 3 weeks, pulling the project together."
"At the end of all of this, the CEO said he would pay me $30,000 or give me 30,000 pre-IPO company shares. Of course, I said I would rather have the cash. I have never heard of these guys before. I invoiced them, finally got the cash, and waited nervously for the check to clear. You know what the name of the company was?"
"Broadcom."
"If I'd taken those pre-IPO Broadcom shares - man, I would have been set for life!"
"So running ESNUG has been a great marketing tool for me, and a great research tool for everyone else in the industry."
"Meanwhile, back when I was just doing ESNUG on my own, a company called Genedax called me and said they wanted to put my stuff on a web page. It was at the beginning of the dot.com silliness. I needed a name, but couldn't call it Deep Throat - that was already taken - so I came up with DeepChip. 'You're in DeepChip now!'"
"Eventually, Genedax went belly up, and later EETimes picked up ESNUG and that's where the money comes from now. I have a strict separation of church and state agreement with EETimes. They sell the advertising on DeepChip, but I have complete control over the editorial content. Everybody advertises on DeepChip - Cadence, Mentor, Magma, Synopsys, Prolific, Summit Design - it goes right down the list of EDA vendors - which makes sense because DeepChip gets an average of 26,000 visits and 65,000 page views every month. In addition, despite the ESNUG name, people are constantly using the DeepChip search engine to find the dirt on lots of non-Synopsys EDA tools. I noticed last month there were searches on Cadence, Mentor, Magma, Nassda, Verisity, Verplex, Monterey, LogicVision, Atrenta, Sequence, Real Intent, Pulsic, 0-in, and Forte. ESNUG is obviously much more than a Synopsys-only tool discussion."
"I think of EETimes as the legitimate press and ESNUG as the illegitimate press. In most cities in the U.S., you have the legitimate newspapers with its legitimate news. And then you have the underground newspapers with the nightclubs, the dirty districts, and the naughty personal ads. EETimes wants to have both, so they sponsor DeepChip. They like it because they can sell ads in both markets. It's a clever business model actually."
"In running ESNUG, my idea is to always keep it very user driven, and that means allowing 'true' anonymous postings. Sometimes the only way the users can share the truth is from behind a mask - but I don't let people abuse this privilege. For instance, if I get a Yahoo e-mail, I don't trust it. To find out if it's a real user, I insist that they send the e-mail again from their real account at work. I've caught Synopsys, Cadence, and just recently Magma sending me e-mail from a Yahoo account posing as a customer. Most of the smaller EDA companies try it at one time or another, too. It's fun to catch them."
"Anyway, anon postings may be in the past for some users. EDA customers who used to be anonymous on ESNUG are becoming non-anonymous by their own choice. After all, you might get laid off next week. All the relationships you've fostered inside your company don't amount to squat once you've been kicked out. Nobody gives a hoot about you or your talents - all that's left is your reputation outside of your current employer, so some of the smarter engineers seek to be publicly known on ESNUG now."
"I've seen an engineer get laid off, go through the shock of getting laid off, and then some time later I'll see a flurry of activity with that engineer's name in the DeepChip search engine. The people who are interviewing the laid-off engineer for a new job are checking him out! They're reading his stuff in ESNUG to see if he really knows his stuff. It's kind of an employment qualification check: 'Does this guy really know something, or is he just pretending?'"
"I'm also seeing a new category of people on ESNUG. They post their names, say their ex-company name, and say they're looking. Recently an ex-Agere guy said he was looking and then the Magma people phoned me wanting to interview him."
"Sometimes, I've even provided direct job references for people. They may have been using an anonymous name in ESNUG - like Kenny from SouthPark, for example - and if an employer contacts me about him, I verify the anonymous name is his and show the prospective employer that he's good by actually reading his posting record. I'll go into the DeepChip search engine, type in the name, and boom - there's intelligent, smart, with-it stuff. I've personally helped approximately 20 engineers get jobs this way. The funny thing was that it was their own ESNUG letters that impressed the employers; all I did was tell the employer, 'Hey, look at this, and this, and this on DeepChip, and you'll see Mr. X really does know his PrimeTime and Astro quite well.'"
"It's funny how ESNUG and DeepChip have grown way beyond their angry origins."
"Anyway, on the personal side, things are changing on my farm. The landlord got rid of the sheep, cut up the land in back, and zoned it for development. There's a little bit of sadness there. Where I used to see sheep, I now see kids with motorcycles and dirt bikes riding around on the construction site."
"But I'll probably never move to Silicon Valley. Being an engineer in
Boston isn't a bad thing because there are also doctors, lawyers, and
teachers here. Engineers are respected here socially. In Silicon
Valley, everyone and their grandmother is an engineer - and they're
101% male - which makes for a nightmare of a dating life. So no,
Silicon Valley isn't for me as a single guy, thank you."
(Editor's Note: John Cooley's e-mail address is jcooley@TheWorld.com for those interested in joining ESNUG. The DeepChip web site is at http://www.DeepChip.com.)
|
