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.
Carbon’s Hal Conklin: A conversational random walk
January 22nd, 2013 by Peggy Aycinena
Several weeks into his new gig at Carbon Design Systems, it was a pleasure to speak by phone with Hal Conklin, VP of Sales and Marketing at the company. Conversationally, it was a bit of a random walk.
WWJD: Prior to joining Carbon, what were you doing?
Hal Conklin: I was doing enterprise software development for the government. I enjoyed the work, but not working for the government. I did that for 3 years and was quite successful, but it wasn’t something I wanted to do long term.
WWJD: So what inspired you to join Carbon?
Hal Conklin: I’ve been in EDA for quite a while and know how important it is to be with a company that has a product and customers who renew [their licenses]. Carbon is particularly important today with the changes that are coming in virtual prototyping. So given all the good stuff I’ve been seeing, I joined the company.
WWJD: What are those changes?
Hal Conklin: With ARM processors going to 64-bits and higher-performance fabrics, you can no longer put vectors [on chips] and test them the old-fashioned way. You have to use FPGAs, emulators or Carbon tools. But now the software solutions are [better than] the hardware solutions.
WWJD: At this point in time, who would disagree?
Hal Conklin: People accustomed to doing hardware prototyping and emulation would argue there is always a place for hardware prototyping, but in scaling to more users the cost of those hardware systems become prohibitive.
WWJD: If that’s the case, why did Synopsys buy EVE?
Hal Conklin: I can’t say why Synopsys [made that decision], but one of the holes in the Synopsys offering is accuracy. What they’re hoping to do, perhaps, is to use EVE as a prototyping tool to increase their accuracy.
Our customers say if the models they’re using are not cycle-accurate, the prototyping platform and models, when they look at the software and architectural tradeoffs – the corner cases – they’re just guessing. But they can’t afford to continue to wait for hardware [validation].
WWJD: So who’s the competition for Carbon?
Hal Conklin: In the grand scheme of things, it’s probably Synopsys. They’ve purchased CoWare, obviously an important ESL player, and virtual prototyping is a natural extension of that business – although the customers they’re selling to might not see it that way.
WWJD: Why wouldn’t your customers see it that way, as well?
Hal Conklin: Our customers understand that if you’re not cycle-accurate, you can’t use cycle-accurate models and Synopsys can’t offer that.
WWJD: I often hear people suggest that Synopsys and Cadence make it hard for startups. Would you also level the same complaint against Mentor?
Hal Conklin: Sure, they’re wholly focused on the broad platform. What Carbon has to do, instead, is to be laser-focused on a single problem.
WWJD: Sometimes talking to EDA folks, I think: Why don’t the customers stop putting themselves at the mercy of the EDA industry, and just build and maintain their own tools?
Hal Conklin: It’s always a make-versus-buy decision, and only a few companies in the world – Intel and IBM, for instance – have the size to do that. Samsung’s expertise, for example, is not CAD tools, but making products. [They see EDA tools] as a commodity that they can get from anybody, and not something they would want [to develop and support internally]. That’s why they made their strategic investment in Carbon.
WWJD: But wouldn’t that also be true for Intel and IBM?
Hal Conklin: Back when I was at Cadence, Intel had it’s own layout tools. Back then, it actually took us 4 years to convince them that our tools were better.
WWJD: Having said that, where will the EDA industry be 20 years from now?
Hal Conklin: Obviously, Moore’s Law will keep going and there will still be a need for EDA. It will probably be a more mature [industry], but you’ll still be getting all of your solutions from the EDA companies.
WWJD: And how many EDA companies will there be at that point?
Hal Conklin: Two or three Big Guys, pretty similar to what we see today. [Of course], EDA will keep changing because the problems will keep changing. But just as it’s always been hard for the Big Guys to develop solutions, it will still be [relatively] easy for them to acquire solutions. So even if we see fewer emerging technology companies in the future, the need for startups in EDA [will continue to exist].
WWJD: When you say it’s easy to acquire solutions, isn’t it still hard for the Big Guys to integrate those solutions into their companies?
Hal Conklin: Yes and no. Typically, if you don’t have a competing product inside your portfolio already, companies like Cadence, Synopsys, and Mentor have shown they can acquire complementary technology. But when they’ve already got a competing product [in house], it is hard to do.
WWJD: And is the problem integrating the personnel, or the product?
Hal Conklin: It’s both the personnel and the product. With software vendors, the software dies pretty quickly without the people [who developed it]. Although, lately the Big Guys have been doing [a better job] of acquiring and keeping the people and their technology.
WWJD: Getting back to Carbon’s expertise, why is virtual prototyping still described as an emerging technology?
Hal Conklin: For two reasons: making an accurate model and making it run faster. If people could simulate the RTL and boot their software on it, that would be the best virtual prototyping because the simulated hardware could be used [as a platform for the software]. But the fact of the matter is, it’s still too slow.
Nonetheless, our approach at Carbon is a software approach, and we believe over the next 10 to 15 months we will deliver products [that meet the need].
WWJD: What’s going to happen in the next 10 to 15 months?
Hal Conklin: We have agreements with Samsung and other companies to help speed the virtual prototyping problems. Specifically, Samsung’s recent investment in Carbon is letting us leverage our expertise and work to our next-generation prototyping product. Although, the actual product offering hasn’t been announced yet.
Of course, there’s always competition. If we look in the virtual prototyping space today – there’s hardware emulation, hardware emulation as a software product, there are FPGAs, and there are several other competing approaches. Only time will tell which is the best approach, but right now companies are looking to solve their problems anyway they can.
WWJD: So with virtual prototyping still sounding like an ’emerging’ technology, has there actually been any progress over the last 5 to 8 years?
Hal Conklin: Yes, there has been tremendous progress! But when you’re getting to things like driver development and operating systems on mobile devices, when you demand such an increase in speeds, it’s always a catchup game. That’s the nature of EDA: Just about the time you’ve got the problem solved, somebody doubles the size of the problem.
WWJD: What about DAC these days?
Hal Conklin: DAC used to be a very important conference, but now it’s harder to get to the customer base directly. So our approach has been very successful – to team up with ARM to show customers what [our solutions can do].
WWJD: Then why not have ARM as an investor?
Hal Conklin: They are an investor! We bought their tool set from them. That’s been very strategic on both sides. They’re modeling strategy is our modeling strategy. We can distribute the ARM model and really help them out in the marketplace. From a synergy standpoint, it’s obviously a [win-win]!
WWJD: What kind of tablet do you carry?
Hal Conklin: I have an iPhone, but I don’t carry a tablet – I use a laptop. If I were just doing email, I [might be interested], but I do a lot of software development, particularly when I’m on a plane, and for that I need a computer.
WWJD: Do you do software development for Carbon?
Hal Conklin: I do some web development [outside of work], but from a company point of view, it would never be my forte. For most of us, it’s an easy decision: The executive team focuses on what they’re good at. We know if you try to go to places you don’t belong, there are always problems!
“Hal Conklin has a long and distinguished career in increasingly responsible roles. Most recently, he was CEO of BlueSteel Solutions, an IT management and consulting company he founded. Previously, he co-founded EDA startup CLK Design Automation and served as vice president of sales and marketing.
“He has also served as vice president of sales for Revit Technology, a company acquired by Autodesk and vice president of worldwide sales at EDA software supplier Chrysalis Symbolic Design, now part of Synopsys Inc. Prior to Chrysalis, Conklin was a top sales performer at Cadence Design System, Valid Logic and Analog Design Tools.”