EDA is not going to go away – neither the technology nor the business of the technology is going to cease. If you follow the polemics and pessimism that play themselves out on the Yahoo financials, if you listen to the folks who decry the monopolies of the large players in EDA, or the lack of qualified grad students, or the inability of the industry to look beyond Moore, you might believe the apocalypse is upon us.
But it simply isn’t true.
All you have to do is spend some time with those who drive the technology, those who push the envelope forward day by day – and as much as I would like to believe it’s the academics who do the driving, above the fray and raw fury of commerce – it’s actually those in industry who embrace the optimism of the future. Because, they’re not throwing in the towel. They believe in this stuff. Having achieved greatness in the past, they believe they can accomplish even greater things going forward.
Laying a billion transistors down in a design, etching that design into various layers of modified sand and metal – this stuff is not only cool and unbelievable, it’s supremely stimulating to those involved in its development, drives the global economy, and offers the promise of massive improvements to quality of life for those who occupy that globe.
2009 & The Audacity of Design
We’ve all heard that 2009 will be the year of change, but it will be far more than that. I thoroughly believe that next year, EDA will pick up where it left off a year ago. Surely next year, we’ll get back to obsessing over the technology, and not over the tumult of tanking stock valuations.
I know this to be true because I just returned from edaForum08 in Dresden, Germany. While there, I had the chance to listen and speak with numerous participants in the growing EDA ecosystem developing in Germany, thanks in part to infrastructure investments from the government. Believe it or not, the German government actually knows how to spell EDA, and even more astonishingly, believes it’s important to provide seed funding to foster regional design automation capabilities to support companies that need such tools.
Folks on the panel I moderated in Dresden – Peter van Staa from Robert Bosch, Christoph Heer from Infineon, Thomas Hoetzel from ZMD, and Gerd Teepe from AMD Saxony – emphasized their commitment to obtaining (a significant portion of) their design tools from third-party vendors. Other folks on the panel – Mentor Graphic’s Mark Croft and Synopsys’ Gregor Wiethaler – were equally committed to ongoing collaborations with their customers. It was good news all around for the EDA ecosystem in Germany, and around the world, and should provide additional reason for optimism across the industry.
[To read more about edaForum08, please check out “edaForum08” on EDA Confidential.]
Meanwhile, just as compelling as the sessions in Dresden was a recent phone call with Cadence’s Dave Desharnais. Although an 80-percent decline in CDNS over the last year has captured the attention of those who follow EDA (and the even greater decline in LAVA), Desharnais was totally upbeat. He’s delighted with the latest Cadence tool offering he’s helping to promote and has no doubts the company has a bright future ahead. Who gave him the right to exude optimism? He doesn’t need anybody’s permission. He thinks Cadence still has a role to play in the future of the semiconductor industry, and believes others at the company support him in that optimism.
Add to that, the fact that Tela Innovations just announced an additional round of financing from Cadence, KLA-Tencor’s investment unit, and Qualcomm to the tune of $5.5 million and you’ll see why there may be light at the end of the tunnel. Ergo, I say vote for The Audacity of Optimization and Automation. Vote for 2009 and the Audacity of Design.
Jim Hogan – Optimist and Realist
Before leaving for Europe, I spoke by phone with Jim Hogan. One of the most affable men in EDA, and one of the few ‘household names’ in the industry able to engage in a candid discussion without PR counsel on the call, we had a lively conversation which I thoroughly enjoyed. At my request, Jim reminisced about his early career spent establishing devices physics labs for both National Semiconductor and Philips.
Hogan chuckled: “That was a long time ago! Back in those days, we used to build things overseas and we didn’t have email, so device physics stuff needed to be available locally. I had the good fortune, at a relatively young age, to put labs together around the world for National and Philips. I had the chance to deal with multiple cultures first hand, in Asia and Europe – Indonesia, Malaysia, Singapore, Thailand, Korea, Taiwan, Japan, Scotland, Germany and Israel.
“Of those places, I spent the most time in Singapore and Malaysia, then Scotland and Munich, plus a year in Japan. Experiencing all of those cultures was a great opportunity for me.”
Hogan laughed and added, “Of course, in those days nobody was flying first class. I think I paid my dues by flying around for years in coach!”
After Philips and National, Hogan did a stint as COO at Smart Machines, where he managed the transition from 200-mm to 300-mm wafers. Hogan said, “When we went to 300-millimeter wafer manufacturing, things were a lot heavier, so we took the humans out of the transport system [on the shop floor]. Also, the 180 and 250 nanometer process geometry production processes were vacuum based, so we had to build transport mechanisms that could bear the weight of 300-millimeter wafers, plus survive 3 million cycles. It was definitely a tough engineering problem!
“To solve it, we built robots in cluster chambers and sold them to KLA and LAM, and they put them into their cluster machines. In fact, that technology is still being used today by Applied Materials. Plus, there’s an additional back story here. Part of [Cadence co-founder] Jim Solomon’s virtual reality effort, Zulu, builds moveable stations on 5 axes to give people more realistic opportunities to experience rides. We were still a ways away from making that functionality viable back in 1998 [when we worked together at Smart Machines], so we decided to go into the robot business for semiconductors instead, which was a natural path for me.”
After Smart Machines, Hogan moved to Cadence where he served in a variety of roles over the years. His favorite? Hogan replied, “Working in the field with the AEs and the customers. Every day, we shared the problems and challenges of pushing the technology on the semiconductor side, with [what was then] less mature software. We worked in real time to get it done, no matter what.
“AEs are really smart guys – they’re not sales guys, they just want to make their customers successful. It’s true they get paid well, but not as well as the sales guys. In my opinion, the best sales guys are actually former AEs. I really don’t want to offend anybody, of course, but the AEs really put their customers first. They’re the ambassadors between the customer and the factory. Working as an AE, or working with the AEs, is a good place to be.”
Hogan described his other roles at Cadence: “Over the years, I was Senior Vice President for Business Development in the office of the CTO, served as a Cadence Executive Fellow – that was a great job! – and then was President of Cadence Japan.”
I had thought the Japanese subsidiaries of EDA companies always picked a Japanese national to head the organization. Hogan agreed, “If you can find someone there to fill the role – yeah, it’s reserved for a Japanese national. But our President of Cadence Japan went on to run SAP in Japan. He was a really smart guy, had gone to Harvard, and was very hard to replace. At the time, I was a Cadence Fellow and [Cadence CEO] Ray Bingham asked me to go to Japan and take over. I was there for about 18 months, off and on, and then for a solid year I was there every other week. At the time, Japan was fully one third of our market, and that’s what it took [to keep things going] until we found someone to replace me.”
He laughed again, “Not surprisingly, I knew exactly where the springs were on seat 2A between San Francisco and Tokyo. In fact, I’m still traveling off of those miles!”
After Japan, Hogan served as Corporate VP of Marketing and Corporate VP for Field Operations (working with those AEs), before leaving Cadence to become Senior Vice President of Business Development at Artisan Components. What prompted the move?
Hogan responded, “How did I end up at Artisan? Well, there comes a time in your career at any company … Ray and I agreed I needed a break, a sabbatical really. That’s how I got to have my first 2-week vacation in my life. My kids were all home at that point. Everybody was spending the summer on our farm, we had adopted a kid, and we had had our first grandchild. We’ve got four kids, one grandkid, plus a couple kids we’ve ‘adopted’ that hang out at our house, so we’ve always had a lot of animals and a lot of kids. My wife’s a really great gal – I actually live vicariously through her! – so that summer we just had a great time.
“I loved hanging out that with the kids, who then were in their early twenties. Believe it or not, we were spending $300 on groceries every couple of days! Anyway, at that time, I was serving on the Board of Nurlogic in San Diego and got chartered to try and sell the company. I ended up selling it to Artisan. After the handshake on the deal, [Artisan CEO] Mark Templeton told me the story of the company. It was clear to me Artisan wasn’t amplified enough in the investment community, so I gave Mark some advice in that area. He asked me to help them out, it sounded interesting, and that’s how I ended up at Artisan.”
Remembering the famous/infamous Artisan strategy of giving things away for free, I asked Hogan if he had been the architect of that mindset at the company. He said, “No, that was before my time. Mark Templeton and Lucio Lanza came up with that. The thing is, you could use the Artisan product [libraries] and license it, and then the foundries paid the company for that use. It turns out that royalties are a high-margin business – the only costs are the audit costs, which run about 5 percent. So, if you can get a royalty business going, you’ve got a real cash-infusion company. We were so successful, we were actually the first tech offering to go public after 9/11. Later, Mark and Lucio sold Artisan to ARM for a pretty good premium, which definitely helped pay for the groceries!"
After Artisan? Hogan said, “My birthday’s in late November and that’s always a time of great reflection for me, a chance to look at what’s happened over the past year and what may happen in the next. [After the sale], I felt my job was done at Artisan and it was time to do other things. I tendered my resignation and looked at a couple of different opportunities. [At that time] Ray Bingham and Don Lucas asked me if I would consider coming to Telos Venture Partners [an early stage investment organization started by Cadence].
“I’d never been on the VC side of things before, and it seemed Don, Ray, and Bruce Bourbon would be good guys to work with. Cadence put $50 million into the thing [to start with] and invested in new companies. One was Clear Shape, which was bought by Cadence, and one was Ponte, which was bought by Mentor Graphics.
“The way it works in venture capital – if you have one or two exits, it pays for everything. So the Clear Shape and Ponte deals were those two exits. Alongside those [events], I was also fortunate enough to have been an independent investor in Brion, a company with a really interesting DFM story. [Brion was purchased by ASML in 2006.] They were part of the move to consider manufacturing variation in design, which had also [provided the motivation for] Ponte and Clear Shape. So all three – Brion, Clear Shape and Ponte – were bought. I’m pretty proud of that. Also, Xpedion was bought by Agilent, so everything I was responsible for at Telos was bought. It may have been beginner’s luck, but I still feel pretty good about that record.”
Speaking of records, we’re all aware the 2008 earnings for Cadence are pulling up a ‘tad’ short of expectations. I asked Jim if he would candidly evaluate the future of the company.
He answered carefully, “For whatever reason, partly because of their own recognition of revenue, but also because they serve an industry which is down – semiconductors are being consumed at a different rate these days, they’re being consumed more, but not at the same revenue rate – so how do I see Cadence? Well, standing outside of the company as an investor, I see they’ve got all that free cash, [which translates] into a heck of an opportunity. If you could get some of the costs down, plus make your customers happy, you could make a lot of money at the current stock price. It’s all a matter of execution.”
I reminded Jim that the last time he suggested Cadence was a good investment was in a brief conversation we had in the Santa Clara Convention Center on September 10th, when CDNS was selling at $7.50 a share. At the time of our phone call on December 2nd, however, CDNS was selling at less than half that price, and this week has even dipped below $3.
Jim chuckled, “Yeah, I’m not being a Pollyanna here, but look – there are some core assets in the company and you’re just not going to displace Cadence in the market [given] their positions in place and route, custom design, and verification. At the core of the EDA market they’re still a very viable entity. In contrast to Magma, for instance, you see a much broader portfolio with Cadence.
“So what do you do to try to fix things? You continue to try to invest in the technology and you work to satisfy your customers. When the customers invest more, Cadence will make more money, and in a couple of years you’ll see a pretty strong performance once again. From my point of view also, a healthy ecosystem in EDA is good for everybody. [That’s why] I would like to see a strong Synopsys, a strong Mentor, and a strong Cadence. Frankly, however, the market’s not big enough for four [big players]. It may not even be big enough for three, but [if we move in the right direction], EDA will be able to get some value back.”
I asked Jim if Cadence owns some of the blame for the devaluation of the tools through their widely reported tool bundling strategies. He replied, “All [of the vendors are] guilty of commoditizing things. Logic verification, for instance, is pretty commoditized. The only thing that saves it is that there’s an enormous amount of simulation required today, so the market is not convinced it’s as commoditized as it could be.
“Meanwhile, in the classic EDA world, logic verification and functional verification will continue to be the number one problems for designers, handing off the design to manufacturing is still going to be challenged with DFM effects – second-order effects that impact yield – and place and route will continue to be challenged by Moore’s law, so the need for a strong router will always be there.
“The reality is, there will not be that many leading-edge designers in the world going forward, but there will be lots of customers behind the leading edge. So, when you ask me if Cadence contributes to the problem [by devaluing the tools], it’s an unintended consequence of going after market share that results in a bundling effort that [effectively] commoditizes the products.”
“If you step back far enough, EDA is actually an IT function, and look what’s happened in IT. The industry has responded to a variety of forces – Sarbannes-Oaxley for example. It’s an enterprise function managed by big companies and implemented by guys like Accenture; a solution is created for non value-added things.
“You have to ask – is semiconductor design necessary? Is it part and parcel of a semiconductor company any more? Or, is it the case that semiconductor companies have got to only run software [as their core business]. Look at Nokia, a great example. They sold their semiconductor unit off to STMicro. Now Nokia does the phone function and the power requirement, ST and TI build the chips, and Flextronix builds the phones. Increasingly, the system company has a set of [partners] that helps them solve the problems of product development. However, the question remains: Can you maintain 60-percent gross margins, because you have got to make those kinds of margins to attract additional investment.”
Based on his comments, I asked Jim if there could be any sort of revival for EDA in the immediate future. He said, “I’m not sure the ‘E’ matters any more, so maybe a revival is not what matters. What matters is how do we serve the software guys? How do we contrast FPGAs versus ASICs versus structured ASICs for the customers?
“FPGAs have not slowed down at all in the current market, even on the high end. Look at Cisco, for instance. They still have a lot of FPGAs. They wanted something that was easy to upgrade in the field, and would allow them to never throw a piece of software away. They know that if an FPGA can run fast enough, and not use too much power – even a high-end FPGA, even at Cisco prices of $3000 a chip – it’s a great opportunity. Cisco sells software in a metal box, and [in a similar way], Nokia sells software in a plastic box. It’s this opportunity to look at all options which is [the real secret] going forward.”
I advised Jim that I felt compelled to address the rumors that he might be the next CEO at Cadence, a company that at this writing has been without a permanent chief executive for over 2 months. Jim answered, “It’s nice that people think I could do the CEO job, but it’s completely the Board’s choice and they’ve got their own criteria. And, they’ve got to answer to the shareholders.
“Cadence needs clear leadership and direction, and they would benefit from having somebody in the leadership role who knows a little about EDA and a lot about the technology of software. Plus, the new leadership will need to extend to the customers and to the employees. First, however, things need to be stabilized so the company can bounce back. Then there needs to be some predictability in the business cycle, and transparency. I think I could do some of those things, but it is going to require a team. Whoever they pick will need good people around them.
“At the end of the day, good leadership needs to look for divergent views. I believe in divergent views because when you synthesize something from those views, the final [solution is that much] stronger. Running a corporation is not a democracy, but it is a place where people should be able to voice their opinions. And, once a course of action has everybody on board, there shouldn’t be a pocket veto on the plan.
“Alternatively, when there is a leadership vacuum, everything becomes very political, which is one reason I really love startups. They’re simple by contrast, compared to a big company. You eat, work, and die by the success of the company – of course there’s always politics, even if there are only 2 people in the company.”
I asked Jim if EDA could grow to be something bigger in the organization, something akin to an enterprise application that linked design to the economics of doing business. He noted my model was not the only one that would promote growth in the industry: “Synopsys, for instance, has some really excellent IP offerings. Plus, they bought Synplicity and now have an excellent FPGA offering, and they are working toward an excellent system design offering. [Of course], a good EDA company also has to have a great services offering.
“To grow the EDA industry, you’ve got to have a good ecosystem, you’ve got to invest in partner programs, plus [attend to] IP as a key ingredient in the mix. And, if there is going to be growth through M&A – my experience at Cadence taught me that it has to be done purposefully and in the proper sequence. A strong EDA company has to be in every digital flow, [attend to] signal integrity, and have [strength in] place and route. So, I agree with your model to the extent that an EDA company can try to be come part of the larger entity, particularly if the [services component] of the business becomes more of a practice, like Accenture, than a service,
“Historically, one of the attractive things about EDA was that software provided excellent margins and good revenue quickly. But that was before the trouncing in 2008 of the financial community. Will the investment community wake up [once again] and realize the opportunities in EDA? Of course, they will. We’ve seen these cycles before.
“My mom was a process engineer, a heavy duty physics person. She was laid off 4 times in one year in the mid 1970’s. When I bought my first house, I was paying 15% interest. We’ve seen these things before, but I’m old enough to know it won’t last forever. We’ll work our way out of this. I know we can tough it out and grow again.”