What Would Joe Do?
Peggy Aycinena is a freelance journalist and Editor of EDA Confidential at www.aycinena.com. She can be reached at peggy at aycinena dot com.
Joe Costello: Orb, Oasys, Epicenter
May 16th, 2012 by Peggy Aycinena
On May 1st, Joe Costello was standing in his office at Orb Networks on the 6th floor of a building in downtown Oakland. When we started our phone call, he said, “I’m looking down on Broadway and there’s a massive march out there. It’s crazy — wish I could send you the video!”
It was, of course, the May Day Occupy Oakland march, which seemed just about right for this long-planned interview.
Twenty years ago, Joe Costello was CEO at Cadence; today he’s President & CEO at Orb Networks, a company that’s “cranking away at cool stuff in the media space.” Twenty years ago, Joe Costello was the epicenter of EDA; today he’s roiling things up elsewhere in the technology ecosystem.
So first we talked about Joe’s present and future, and then we got around to EDA’s present and future and What Would Joe Do if he was back in the epicenter today.
First off, what cool stuff does Orb do?
Joe said, “Apple has something called AirPlay, which is a nice feature for moving media for people who have all Apple stuff. What we do is AllPlay, which does the same thing for everybody else.
“Orb does two things; we build products and demo-vehicles for the world, Orb Music and Orb TV. But we don’t think Orb will be the ultimate consumer of our products. The key is the underlying architecture and platform. We license the AllPlay platform, so our customers can build all those cool Apple-like functions into their own products.
“Our first OEM product is set to launch in a month, a set of speakers for the home built by a well-known manufacturer that’s building our technology right into their product. In fact, from Android to Apple, we work on everybody’s stuff – anything you can control with a pad, pod, or phone.
“How did we end up with our offices here in Oakland? It was a natural sequence of events. The company spun out in 2004 from an incubator that I started in 2000, and when we looked at where to put the company we knew we had a lot of talent between San Francisco, Berkeley, and Oakland.
“Initially we were housed in Emeryville when the real estate market was really terrible. But now Emeryville’s the hottest property in the East Bay, so we did the normal thing. We got out the map and used pins to locate all of our engineers – the company’s mainly engineers. It was clear we had to be on a BART line, and that made Oakland a natural choice because there are two different lines the meet here.”
“So this is my full time day job these days, right here in downtown Oakland.”
Joe may have a full time job, but he’s got other things going on as well. He’s Chairman of the Board of Oasys Design Systems, a company focusing on a “new platform for Logic design called Chips Synthesis.”
How long has he been involved?
“Three or four years,” Joe said.
“As you know, after Cadence I wasn’t doing anything in EDA, I was out of EDA completely. But then it felt like there was a paradigm shift coming, felt like there was going to be a need for new technologies and tools to handle that shift. So when Paul van Besouw [Oasys President & CEO] came to talk to me, it sounded like we were getting even closer to that new era and it seemed like the right time to visit changing the rules.
“I actually didn’t instantly go and invest in Oasys or join the board. Instead, I made suggestions for this and for that, to prove the company’s technology was real. But after a year Paul came to me said, ‘I’ve done it all, everything you said I had to do.’”
Joe laughed, “I knew then I had to make good on my word, and that’s how I got involved.”
Joe added, “I do believe it’s a different time now. The dynamic in EDA is so different than when I was first in it.
“It’s hard when you build a company at a certain time in your life. Your reflexes still send you back to that time. Things are not moving as quickly as they were when I was first at SDA [Solomon Design Automation], and then at Cadence. Back then, there was a much more rapid angle of attack.
“Now EDA is a mature business and, in general, industry moves slower today. The problems are more challenging, it takes more time, and it takes more energy and effort for people to make changes. Look, you see it even on the investment side. EDA was the darling of investors when I got involved in the 1980’s. Today, EDA is a four-letter word.”
I mentioned a comment legendary EDA investor Lucio Lanza made some years back: Don’t brag on your next round of investment, because with each round you’re selling off more of your company and losing that much more control.
Joe responded, “The more significant the challenge you take on, the more money you need to raise. Unfortunately, despite Lucio’s comment, that’s just how things work. If you’re building a little tiny app that only take 6 months and a couple of guys, you can do some viral marketing, spend a little money, and maybe succeed.
“However, if you’re taking on a core technology, if it’s a complex set of technologies you’re building, you have to have staying power to deliver it, to modify it, to get it right, and to stay with the customer while they change mindsets and get used to the new technology.
“Yeah, it would be great if everything was just a cute little app, but VCs today with Lucio’s mindset say: ‘We won’t give you any money until you’re actually selling your stuff.’ That may be wonderful where you’re investing with people who have a proven product, but …”
Joe continued, “The old school thinking in the VC community: Think strategically, look at the direction the market’s heading, and then make a bet on a group of people. There are still blips like that out there today, but it’s only those places where you can invest with far less rigor.”
And if you need to invest with more “rigor” in companies already selling products, I asked, aren’t they in the driver’s seat? Don’t they then get to pick and choose between investors?
Joe said, “You’re right. Today, companies chase putting money in super high valuations. They’re playing a different kind of game, a different kind of investing. These wild investments in companies like Facebook? As you said, do they really need the money at the point they already have proven products?”
Joe added, “Think about it for a second. In the space we’re talking about, in EDA, there’s not really an opportunity today. I’m a big believer in, if you start a company in EDA, the thing you want to do is to develop some useful capability as fast as you can – not even a product per se – to show a customer that there’s value there. That’s the important thing. As fast as you possibly can.
“That’s the only real value in EDA. You need to show how you can relieve some pain for the customer, and give them some advantage. The fastest path [to that solution] is extremely important. And you can’t do that in just a couple months with just a few people. Today’s design challenge and complexity make it impossible.”
But small companies can’t stay afloat long enough to demonstrate a new solution, I noted.
Joe said, “Yes, and the place where I see this? Listen to the venture guys. One of the reasons they’re not investing is because EDA is such a closed marketplace. It’s extremely difficulty to break in, and I’m not the only one who sees it that way.
“The VCs guys will always make a buck wherever there’s an opportunity, but today in EDA there are only two major players in this marketplace and they’ve locked down the design flow. They’re discounting [the point tools] up to 90 percent, and economically making it extraordinarily difficult to put anything from anybody else into the flow.
“If you come to the customer with a new tool, their response is: ‘Ohmygod, the disruption! It could break something, so I want to just leave well enough alone!’
“So this combination – the economics make it difficult to penetrate, and impossible for anybody else besides the major players to make money; it’s hard to get the engineers to use the tools; the major players don’t make it easy to insert other tools into their flow; and the patent and IP wars – all of it make this a very hostile environment [for small companies].
“But, even if you could overcome all of this, how much revenue could you actually generate? At the end of the day, you can’t go public and then you’ve only got two companies who would consider acquiring you. Cadence and Synopsys are doing a great job of fortressing, and the VCs are therefore saying, ‘Ew!’
“Why should they invest when you can’t get the customers to pay for the innovation, or to adopt it?”
I asked Joe: Are you ever wistful about the old days? Do you think there’s no more joy?
Joe said, “It’s true, today is not like the early days, but it’s not a wistful thing. It’s just that it was a great time. We were living in a Brave New World, with total excitement and tremendous amounts of brain power coming into the industry. It was exciting and free-wheeling, with things springing up all over the place.
“That said, however, as technology has gone down its path, there’s still a huge amount at stake in these businesses. And it’s not like the design problem is completely locked up. Moore’s Law guarantees that you have to keep marching on, [particularly] at those times when you get these paradigm shifts.
“One of the leading indicators we see at Oasys? Look at our investors! We have angels and two strategic guys, Xilinx and Intel. It’s no accident that these are not traditional VCs for the exactly reasons we just talked about. Both Xilinx and Intel realize they continue to need great technology development, and they see companies like Oasys meeting that need.”
But don’t the big customers continue to create their own tools internally?
Joe said, “In the old days before the golden days of the design automation business, yes they all built their tools internally. But then the EDA companies came along with their intensity and energy, and the companies said: ‘Hmm. Maybe we should see [if this is a better strategy].’
“They’re still doing some internal tools, it’s true, but they also know that given the amount you have to invest to build the complete solution, it’s better to have people focused.
“Customers aren’t working to their core competency [if they’re building tools]. Software for design is different than doing chip design. That division of labor is still right today, and they continue to pay the big EDA costs.
“Although as the big companies in EDA have consolidated, it’s dawning on the customers that they also need to foster innovation with the smaller EDA guys. Now those customers are saying: ‘I’m going to use this, and I don’t care what Synopsys, Mentor, and Cadence think about it.’
“I’m not saying here that Cadence, Mentor, or Synopsys aren’t doing anything today [in R&D], but the customer base clearly wants to see more diversity. It’s critical to them.
“So as things have gotten more and more constrained within EDA, the customers are wanting more diversity and they’re investing [to make that happen]. These guys have a huge appetite for innovation, and we see it at Oasys where Xilinx and Intel are literally investing to make it happen.
“The large companies have so much at stake in this next round, this next paradigm shift, they’re saying: ‘We just have got to find ways to make that happen!’
“That’s when things can get really exciting, which they generate excitement [in the tool providers]. And as that happens, EDA will begin to bloom again, things will begin to rock and roll again. There are changes going on today that are going to make that happen.”
“I’m not the day-to-day expert I used to be, but as things go to smart mobile devices, this is the hot space today. And it’s all big IP-based development. It’s mobile, low power, and high functionality because everybody wants their pad, pod, or phone with more and more features, functions, and capabilities – CPU, graphics, entertainment – and to do that, people are using these big IP blocks in their designs.
“Those three things are jamming the current design methodologies up against the wall, and they’re all going to break slowly but surely over the next several years. People are at the bleeding edge of design, which opens up the possibilities.
“Yes, it’s such a constrained situation today in the EDA industry, there are not a thousand flowers blooming. But there are still people pounding away at things out there, and even the Big Guys in EDA count on that stuff to [for innovation].
“So, there is energy. There are exciting interesting problems in this space that can’t be solved by the Big Guys. There’s life there and opportunity. And when that’s true, it’s a guarantee that young bright [talent] will come along.”
I asked Joe what he thinks motivates people to work so hard in technology. The money? To corner the power? To conquer the technology?
Joe replied, “It’s all about, where is the money for this or that technology, and the chance to challenge the established market leaders. Dominating the market and the money – those are the motivators.
“But underneath it all, they really love it. The engineering guys really love the technical challenge. They love working on the latest, greatest thing. And it’s magic when it works, when it comes out in the market. Look at the iPad 3 – an impossible feat, but they did it!
“This translates into the EDA side of things, too. Under the covers, it’s impossible to get things done today with the traditional flows. You’re going to have to change things to breakthrough to the 3GHz versions that don’t have a battery life of 2 minutes or burn a hole in your lap. The challenges are so exciting, both from the tool side and the technology side.
“I see it all the time, I see the light in people’s eyes: ‘Ohmygod, there might be an answer to this!’
“They want to be the someone who’s motivated by this, they want to be the one who makes change happen, as opposed to incrementalism. They want to get these chips done and they’re excited about the work. Yes, there’s top-level motivations. Sure it’s got to be about the revenue and the profit, and being the market leader.
“But understand that, underneath all that there’s a ton of joy. We’d be nowhere today if there weren’t engineers who do this stuff with incredible joy!”
Joe Costello was awarded EDA’s Phil Kaufman Award in 2004. You can read the presentation speech here, delivered by the late Richard Newton, Dean of Engineering at U.C. Berkeley.
One Response to “Joe Costello: Orb, Oasys, Epicenter”