工程师的故事

haichao

On a frenetic Monday morning...
With the surreal mix of war talk, jittery markets, and the kind of perky chatter that always accompanies the start of the conference season - DesignCon, DVCon, DATE, PCB West, ISQED, etc. - can we just turn away from the frantic headlines and glib EDA messaging for a minute to talk to an in-the-trenches designer about the challenges of his work? Maybe it will help lower the blood pressure a bit.
James Lee is an engineer's engineer - a computer nerd, an avid ham radio operator, and an affably blunt businessman. He's also a walking encyclopedia on Verilog and comes with several decades' experience in ASIC design and EDA product development. And, despite a dreary economy, he's chosen this moment to launch his own company with the help of several others - The ASIC Group.
Lee says, “We've all been having fun with design services for quite a while, and we decided now would be a good time to do it on our own. Besides, downturn start-ups are often the most successful ones - because it's what you do during the downturn that determines how you'll do in good times.”
Lee earned a BSEE from Massachusetts' WPI, worked for a number of years as a memory designer for Data General, and was lured to California by an offer from Gateway. “I was fairly young and decided it was my chance to see the world. So I came to California and started at Gateway, where I was an FAE for a year. Then I was called back to the East Coast by the company and asked to design the first hardware accelerator for Verilog, the XLP. That's a tiny footnote in EDA history - only 10 units were actually shipped. There was nothing wrong with the product; it was just a case of a software company not knowing how to market a hardware product."
“So I left Gateway and my wife and I decided to return to California, where I joined Cadence. I was there until 1996, working with Alan Naumann and the Systems Division when they were moving beyond the IC area and more into the front-end. It was in the early 90's, prior to the Valid merger, and it was lots of fun to work at Cadence. It was probably one of those experiences that really showed the difference between East and West Coast companies. At the time under Joe Costello, Cadence had 900 employees, but had a very youthful atmosphere. Back on the East Coast, Gateway only had 100 or so employees, but was much more conservative.”
After many years with the company, Lee left Cadence in 1996: “The last thing I was doing at Cadence was helping to create market penetration for their Synergy Logic Synthesis product. I was Director for that product. But at some point, the Spectrum Services Group decided they liked Ambit better than Cadence Synergy, and that spelled the end of the product line. I could have found other work within Cadence, but I had been talking with Seva Technologies - Yatin Trivedi and Larry Saunders - about when it would make sense to go back into design services. So I left Cadence and joined Seva, which had about 10 employees at the time. We developed a lot of really good customers - some of which are my customers today with The ASIC Group - and I learned a lot about the consulting services business.”
“Meanwhile, Seva was looking for a way to grow and Intrinsix [in Massachusetts] was looking for a way to expand their West Coast operations. We offered a solid team and the merger was a natural pooling of interests. It all started out great, even heading toward an IPO. Everything looked good, but the combination of 9/11 and the economy was too much. It was too tough for Intrinsix/Seva - just like for so many companies.”
So Lee decided to go out on his own, taking his technical expertise as well as the business savvy he had acquired while working in the Intrinsix/Seva organization: “I knew by the time I started my own company that there are some good customers and some bad customers out there. You need to know how to figure out which customers are hard to work with or have financial difficulties. You have to be very cautious about small companies and understand their credit and cash flow before you start working with them - it's just standard practice to check a customer's credit reports and the banks. I've been doing this long enough that I can get a feel for which companies to do business with - essentially there's a good vibe there.”
Lee is delighted he's continuing to work with customers he interacted with while at Intrinsix/Seva: “SnapTrack is a real software success story. They've got a number of really critical patents in GPS technology and a number of sharp people who are a lot of fun to work with. The SoC chip we did with them, and with Samsung, was on the first page of the San Jose Mercury News a couple years ago. Then Qualcomm bought them and we're still working with the company, and with Qualcomm. And Qualcomm is also just a very good, nice group of people to work with.”
So with some solid customer relationships in hand, The ASIC Group is moving forward. Lee says, “We've got an impressive core of employees with 15 or more years' experience who've worked with industry giants and with little start-ups with interesting technology. We work at the customer's site or sometimes in our home offices via our virtual data center, and we're pretty much like other design services companies. We know that the most critical thing in doing a successful design for our customers is to understand the key factors in their business model - more than just the technical details of what they're trying to accomplish. ”
“For example, if your customer's going to build a consumer device, the ultimate driving factor is low-cost manufacturing in mass production. So you want to help them produce something that may be part of a family of products to help amortize the engineering and production costs over the life span of the product. Also, depending on the trade segment they're in, you may need to help them schedule the product release around CES or COMDEX, or to be in full production by the middle of summer if the customer wants to hit the Christmas market.”
“Additionally, if it's a consumer products company with all capabilities in-house except, say, an ASIC design team, we'll work with their PCB and software teams to help bring an SoC together. If they don't have PCB or SW disciplines, we can help bring those skills into the mix by assembling an embedded software team, for instance, to help them. For us, it's basically a matter of understanding the customer's needs and filling the void for them.”
But Lee emphasizes, “If there's one critical success factor to any project, whether design services or in-house design teams, it's communication between team members, corporate leadership, and the design managers responsible for the project - good communication with people responsible for the next step, the board design, manufacturing, or whatever. All of these teams, in the end, really make up one bigger team. If they are people who want to communicate freely and work openly, then you can see the recipe for success. If there's fighting or bad communication, even the best group of genius engineers can't salvage the project.”
Maybe the world, in general, could take a page from Lee's common-sense approach and straightforward philosophy.

一粒尘埃

遗憾，英文不好！！！

zipl1985

看不懂

motter

楼主水平好高啊！