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.
Hal Barbour: 8 Grand Challenges in IP
May 25th, 2017 by Peggy Aycinena
The first installment in the series, published last week, was a conversation with Sonics co-Founder and CEO Grant Pierce.
Pierce argues that today’s Grand Challenges in IP center around the complexities of delivering sub-systems and related technical expertise to customers, helping develop edge-node devices targeted at Machine Learning, and providing IP for myriad automotive systems – all while meeting demands for greater bandwidth and throughput, and astonishingly low power.
In this week’s installment in the series, Hal Barbour talks about a completely different set of Grand Challenges in IP – those related to the business issues surrounding the industry.
Barbour has some very specific ideas about maximizing the circumstances for both the IP vendor and the customer, but he started the conversation with good news: “We’ve had a fantastic year, with 2-to-3x increase in business over previous years. And more importantly, no single customer represents more than 5 percent of our business.”
Does that fact take on particular meaning in light of recent news that Imagination Technologies has been upended by the threat of losing Apple as its principal customer?
Barbour said, “Absolutely. You’re screwed in the IP business if you depend on just one customer, and Imagination is not the only one. There are others as well tied to the super-smart phone business [who would be compromised] to lose a key customer.”
“It’s true,” he clarified, “we do have a lot of business in the smart phone market, but unlike other people we have a royalty-free business model. We make money through an up-front license fee and not royalties.
“And we do not have a top-heavy customer base. You have much great immunity to the changes that occur in the industry, the down cycles and up cycles, if you do not rely on one or two customers.
“We have always felt that the broad range of our customer base is our great strength.”
“We have been in business now for 23 years,” Barbour noted. “We have a good name and reputation out there, and our products are really competitive.
“We’re also different because we don’t do elaborate marketing or put on lavish seminars. It’s not that we couldn’t do these things, but we keep our costs very low and for that reason can provide a better value proposition to our customers.
“And CAST has really, really good people, many located in less costly areas within the US [than Silicon Valley]. Compared with other vendors who have outsourced a lot of their work to India, we also have remote teams in the Czech Republic, in Poland and Greece. These are very focused groups, tightly coupled to everything we do.”
Barbour acknowledged that it’s not always easy to coordinate across remote teams, but said the company successfully “worked through” those issues a long time ago and is now reaping the benefits.
I asked Barbour if CAST’s success has been due in part to luck, or even dumb luck.
“That’s a good question,” he said, laughing. “It’s been luck and dumb luck, because everyone knows the correct way to start a business is to plan your way, write a business model, and get outside financing.
“But that traditional route has not been us. We don’t see out more than a year, or a year and a half, and we have far more flexibility because we are bootstrapped. We’ve never had a loan and never taken a dime from outside investors.
“And we have no internal financial people whatsoever, and I’ll tell you why.
“When you get money from the outside, the investors demand that you have a vice president of finance. That’s who they’re always closest to in the organization.
“From the investors’ standpoint that may make sense, but from a business point of view, those internal financial guys bring no added value to the company.
“At CAST, we keep our expenses in line,” Barbour emphasized. “We’re not stupid, and it’s not that complicated. It’s about controlling the costs.”
“It’s also about focusing your sales, engineering, and energy towards your customers, and doing it in a fashion that’s good for your customers and good for you.”
“And,” Barbour continued, “we don’t have any commissioned sales people. None.
“Almost every other company has commissioned sales people. Why? Because they say that’s how you get people to work hard for you.
“In the case of highly leveraged sales commissions, it’s done primarily because it’s easy to measure and because you can put a whip on people’s back. But I never believed in putting a whip to people’s back to get high performance.
“When I first started in sales many years ago, I always appreciated the engineers behind the products I was selling. They contributed as much to any deal as I did.
“It’s just a different way to look at the world,” Barbour said.
These ideals seem appropriate in small companies, I said, but what about the large IP vendors in the industry?
Barbour responded: “Over the 30-year history of the industry, clearly there have been lots of acquisitions by Cadence, Synopsys and others, which have had some effect.
“Then there are the large acquisitions by financial institutions, as we’ve just seen with ARM. Our 2-to-3x growth over this last year indicates to us that some customers are going away from ARM because of this SoftBank acquisition.”
Have the costs for ARM products increased due to the acquisition, I asked.
“Trying to find out what you’re competing against is actually difficult,” Barbour replied. “In our case, we’re competing against a much, much lower business model.
“With ARM, for instance, their IoT entry-level model is called Zero and [costs] approximately $40,000. But once you’ve shipped your product based on that core, you’re still looking at another $300,000 or more in royalties.
“Our comparable product has an overall cost to the customer of closer to $100,000 to $125,000 – including [assistance] in customizing blocks – and no royalties, so the cost is far lower.
“And our development engineers are tightly coupled to the sales process, so we can respond to customers’ customization needs.
“The big IP companies charge $3000 to $4000 a day for help with customization. But with us, the cost is much, much less than that. We can provide customized block at less than half the cost of the big IP providers.
“Many times what customers need is not something complex that’s running an OS, or something that’s very expensive. What they need is a deeply embedded processor or an engine that doesn’t require, or justify, a royalty type of engagement. They just want to lower their NRE costs.
“We don’t have a royalty model, so our customers will not be heard saying, ‘This damn royalty is killing us, so let’s go elsewhere – or do it ourselves.’
“This the most important goal,” Barbour said, “to keep the costs of IP low for the customer.”
Asked about an exit for CAST, perhaps an IPO, Barbour was adamant: There are no advantages.
He cited several examples of companies that lost their way after going public, including View Logic: “The CEO of View Logic was Will Herman, a really nice guy and a brilliant engineer.
“We were a partner of his when CAST first got started and every time I went to see them, Will would call me into his office to bounce off some ideas for yet another presentation to Wall Street.
“After the second or third time [we went through that exercise], Will acknowledged he was spending at least 50 percent of his mental capital towards something that had no real bearing on his customers’ well-being or his employees. All of that effort was strictly for the Street.
“For Will, going public was all about feeding the Beast and after a while, the company got so pressured [by the Street], it was sold off – part went to Synopsys and part went to Mentor Graphics.
“We have never had to go down that path, which has continued to give us a lot of independence.
“I can’t tell you we had a long-term vision for CAST when we first started, we just wanted to do logic simulation for models. But it was too late [in the industry] to do a simulator, so we decided to do models instead.
“Had we had outside financing, or were answering to share holders, the financial folks would have asked what the hell we were doing and never allowed us to change our focus.
“I’ve seen some really, really good engineering organizations torn asunder by unrealistic expectations from the financial people. We have not had to [endure that difficulty].
Most importantly, Barbour stressed, hire and/or work with excellent people.
“Our business model is not just about developing CAST owned IP,” he said. “It’s also about working with long-term partners in a tightly coupled business relationship, or serving as exclusive world-wide distributor for people who are experts in their field.
“We work with companies like Ocean Logic, Beyond Semiconductor, Fraunhofer Institute, Sandgate, Silesia Devices, and SoC Solutions. And we have partners in Poland, Slovenia and Greece – they’re all fantastic.
“In effect, we behave like one company, but with each unit operating somewhat like we do – relatively small and managing their costs very carefully. And by operating in geographic areas that are less costly, we can scale much more easily with six or seven development centers beyond our own engineering centers.”
It’s not just great partners but great people who make things a success, according to Barbour.
For instance, he said, “When it was time for me to move from CEO to Chairman, we hired Nikos Zervas, who’s been a great fit. He’s more technically astute than I am, younger, and in all regards better than I am. Everybody respects him as I do. It was a very smooth transition.
“And Paul Lindemann. He’s just fantastic at marketing and has been with us since the very beginning.
“I feel very lucky to have been around these types of really, really good people. Believe me, it’s not me who has made CAST a success. It’s our great people have done that. I’m only one link in the chain.”
Hal Barbour so clearly enjoys his work, I noted he would be working well into his 90’s.
Laughing, he said, “I very much like staying involved in a company I had a major part in building. And I like talking to the customers.
“Why would I want to do anything else?”
Hal Barbour is Chairman of the Board of CAST, Inc. He earned a BSEE and worked several years as a circuit designer before moving on to technical sales and marketing positions with GenRad, Intergraph, and HHB Systems (later acquired by Racal-Redac).
Witnessing both stunning successes and colossal managerial failures, he has applied those lessons to semiconductor IP provider CAST. Pioneering a successful virtual and distributed organization model and using a lean, customer-oriented operations philosophy, Barbour has helped CAST succeed and thrive in a volatile and challenging market.
Tags: ARM, Beyond Semiconductor, Cadence, CAST IP, Fraunhofer Institute, Hal Barbour, Mentor Graphics, Nikos Zervas, Ocean Logic, Paul Lindemann, Sandgate, Silesia Devices, SoC Solutions, SoftBank, Synopsys, View Logic
One Response to “Hal Barbour: 8 Grand Challenges in IP”