Voices: Mansour Izadinia: spreading analog expertise
Mansour Izadinia, senior vice president of IDT's analog and power group, discusses product definition, smartphnones, and opportunities in the analog area.
By Paul Rako, Technical Editor -- EDN, February 18, 2010
IDT (Integrated Device Technology) recently hired Mansour Izadinia as senior vice president of the analog and power group, signaling the company’s growing emphasis on analog- and mixed signal-products. Izadinia has a bachelor’s degree in electrical engineering from the University of California—Los Angeles and a master’s degree in electrical engineering from Santa Clara University (Santa Clara, CA). He has seven patents in the analog field and has authored several articles on the subject. EDN recently interviewed him.
What got you into analog?
I got into analog from school. When I was at UCLA, I always wanted to take the hardest courses. Everyone told me that analog-circuit design was the hardest course, so I tried it.
What was your job path to get to IDT?
I started as an analog-design engineer at National Semiconductor. For a number of years, I worked for [analog engineer] Bob Pease. After 10 years at National, I went to Maxim. I worked at Maxim for 16 years.
What kind of things did you do at National Semiconductor?
I designed many dc/dc converters and motor-drive ICs, such as the LMD18200. I designed a lot of the BCD [bipolar-CMOS-DMOS]-process products. My last year was as a section head for automotive design. I had a group of design engineers who were working on ABS [antilock-braking systems] and air-bag controllers. I designed a lot of custom automotive parts that don’t show up in catalogs.
So you had an affinity toward the processing side of semiconductors, as well?
Even though I’m strong in circuit design, I never viewed an analog-circuit designer as a success unless he had a mastery of process technologies. I always wanted to understand how semiconductor-process technologies worked. I paid a lot of attention to that, and I was as good with circuit design as I was with process technology. I worked on bringing up several generations of BCD technologies at National and at Maxim.
What attracted you to Maxim?
Maxim’s [analog engineer] Dave Fullagar (Reference 1) called me in 1988. I didn’t take the job. Every year, I would get a call from Maxim. In 1994 I felt Maxim was not a start-up anymore, so I went there. My first job was as a senior scientist for portable power. I started to hire people and form a group of design engineers. I designed several Maxim products in the notebook area.
What increased your responsibilities at Maxim?
Back then, Maxim was very strong in portable electronics. I felt that they didn’t have a lot of traction in the communication segment—things like the base stations, servers, routers, and switches. Cisco, Alcatel, and Siemens were not our customers back in 1995. I started championing doing products for this segment. I took business plans in front of our chief executive officer, Jack Gifford. In his classical methodology, he would reject me 10 times, but he eventually bought the idea. If Maxim wanted to be a power-management supplier, we would have to address this huge segment of telecommunication, server, and nonportable applications. The group was a bunch of design engineers and one marketing guy and one or two application engineers. That group grew to about 116 individuals. My last position at Maxim was vice president for the system and power-management group.
Maxim was the first analog company that seemed to organize by market. How did that come about?
This move goes back to the entrepreneurial spirit of Jack. He was an entrepreneur, and he led a lot of people who had the drive, passion, and technical background to explore markets. He believed in engineering excellence. He believed in a business of selling engineering. When Jack saw somebody who was technically good, had passion and drive, and knew what he was talking about, he gave them a lot of rope.
He believed in creating these little fiefdoms—that was his word—that would let these guys all compete for market segments and become experts at those market segments. He believed in technical, product, and application expertise about markets. He also believed that you have to beat your competition by encouraging excellence and doing good execution. He had such a huge emphasis on execution. “I’m going to create these pockets that are run by these driven people who get specialized about end markets,” he thought. I have never met any business semiconductor person who was better qualified.
Was Gifford hard to work for?
In a lot of ways, he was very misunderstood. I think his close friends know what a human person he was. He really did have a soft heart. He held the bar high and drove a lot of people hard, but he had a big heart.
What attracted you to IDT?
I always liked passionate people. Jack was a passionate person. I always wanted to be around people who had drive and passion because I’m the same way. When I was talking to Ted [Tewksbury, IDT’s president and chief executive officer], I could feel the passion in his voice. I have known him for a long time. At Maxim, he had a lot of passion. He impressed me with that.
A lot of people get hung up on creating a 10-year plan that’s all mapped out. I think what’s even more important is that the person is adaptive to change—that he can figure things out. A person with drive is going to figure it out along the way. I like people that are not myopic about things and who are not stuck up about their own ideas. I like people who are open to other people’s ideas and who can build a team. You’re only as good as your team is. Ted understood that. A lot of things that he told me resonated well.
IDT is an established company. Is a start-up culture something you like to build in your team?
Yes. I think the key to success for a large company is to make sure they keep the team’s focus and that they focus on execution. This approach is integral to the success of start-ups. If you look at how start-ups succeed, they get a small group of people who are not bothered by all the other corporate stuff. They execute. The key to success is exactly that. I think IDT is going through this transformation—this focus on execution. Ted started that focus, and I’m carrying it forward. I’m trying to put an emphasis on product execution, product development, and having the right metrics in place. I learned this approach from him.
How do you think product definition should work in an analog-semiconductor company?
You didn’t really need an equipment expert 20 years ago. An op amp is so general-purpose that it would compete on parameters and specs. As systems got more complex and integration took hold, we needed to have people who exactly understand the end-customer system. Sometimes, our customers don’t understand their own subsystems. Many companies don’t have power-management experts, yet power management is critical to the performance of their end products. There lies an opportunity for us to bring that expertise and use it to come up with differentiated products. We alleviate our customers’ headaches. In the future, customers won’t really care what an IC does as long as it solves their entire problem.
|
Apple shows this approach. If you buy an iPhone, it takes you 10 minutes to know how to use it. It’s exactly the same with chips and ICs. If you handed a 300-page data sheet to a customer, they don’t have the time and sometimes not even the expertise to be able to read it. Ease of use also applies to ICs. To do that, we need to have elegance in product definition. We need product definition that’s targeted, that’s specific, and that solves the problem with the least amount of headache. When a customer calls us and wants to use our product, we need to have experts in product definition who can go in and solve the problem in the shortest amount of time. That service is what differentiates our products. Ted has put an emphasis on this issue. We’ve been bringing that end-equipment expertise into IDT.
Some companies call [this person] a product definer. Some companies call him system architect. Some companies call him a marketing person. Some companies call him a field application engineer. It doesn’t matter what title you give that person. You’ve got to have a person who has the system knowledge of your customer and who can bring a value to your customer. No customer ever wants to talk to a salesman or a field-application engineer who doesn’t understand his problems.
So, rather than resent doing your customers’ jobs, you view them as opportunities?
Absolutely. We have to bring value to a meeting. We want to solve this problem that you have. We want to exactly understand that problem. That differentiates companies and drives sales. It’s not just an IC design anymore. It used to be that you would execute on a product specification. If you came up with a lower offset voltage on an op amp, you would win the business. It’s not that way anymore. The world has become so complex and systems have become so complex. The end equipment may have one or two ICs inside. The whole thing is integrated. That guy who understands how to apply that IC is the one who wins the business.
Smartphones seem to have changed everything. They have lots of analog and power-management content. Is that the kind of business you want to do?
At IDT, we have a diverse set of technologies that apply not just to smartphones, but also to e-books and display applications. We have a whole bag of technologies available to us. I think mobile computing is important. I can‘t speculate on whether it’s smartphones, e-books, or other audio and video handheld devices.
IDT also has cellular base-station RF chips.
We have a sizable business in base stations. We understand how the data flows in a base station. We are focusing on some of the front-end technologies for the base station, as well. We want to be a total system provider. We want to provide the entire solution that goes into a base station. We can choose to go after sockets that make sense for us, both in core competency and from a business aspect.
Your knowledge of process seems to play well serving an entire system.
We’re going to be looking at all these system pieces at IDT. In the front end, we look to provide RF devices. On the back end, we are looking at providing the power-amplifier devices. It’s not an issue of whether we need to have analog, digital, DSP capabilities. It’s an issue of providing tools for a complete system solution. I don’t think that any company has a choice in being an analog supplier or a mixed-signal supplier or a digital supplier. You have to have this bag of tricks. You have to be an analog provider. You have to know how to do signal processing. You have to make the best ADCs and DACs. You have to be able to make the best power amplifiers. These are all the little blocks that go into a big system. When Ted came up with this charter of being a solution provider, it meant that the company had to have analog, signal processing, and RF and power amps.
What is your attitude about fabless versus captive fab operation?
We need certain technologies to be differentiated from our competition and to provide special value to a customer. We don’t need to have a fab to have those differentiated technologies. So, whether or not you own a fabrication facility, I think it’s immaterial. We can develop specialized process technologies within any of the captive foundries.
Do you mean real process differences, or do you mean device geometries or IP (intellectual property)?
We have the expertise to do that. I’m a device guy as much as I’m a circuit guy. I don’t need a captive fab to implement those processes. Sometimes, having a fab is a hindrance. A captive fab is targeted toward the entire corporation. Sometimes, they’re so busy with doing what’s right for three-fourths of the company, that, if you’re trying to start a business and you actually need a specialized process, they have no time to give you. Not having a fab is somewhat of a blessing because I don’t have to convince one technology-development group to add a process. I have multiple foundries that are calling us on a weekly basis asking, “What can we do for you?” You have to figure out what you’re trying to do. These days, there is going to be a fab that will implement what you want if you provide them a business case. China is building foundries as if there were no end.
I notice that IDT makes analog switches. Can you comment on that?
I managed the design group for analog switch-product line at Maxim, and it’s a great business. It’s underserved in the marketplace. A lot of companies and a lot of designers think that analog switches are low-tech, old technology. But there are things that you can do with analog switches that you cannot do with more integrated solutions.
So, does it make sense to design a system that is not just a single chip?
Yes. When you integrate more and more of these solutions, what if something goes wrong? What if you get to the end of your product development and you’re trying to ship a product two weeks from now and you didn’t define a certain thing right? That’s when an analog switch comes in to the rescue. That’s how we built a huge business, because errors happen and things change.
I notice IDT has a broad spectrum of part types. Could you comment on that?
Ted has a vision that you’ve got to have these foundations—these pillars that we put in place now to invest for the future. Touch technology is going to be a must. Audio is going to be a must. If you look at what IDT has been doing, we’ve been putting in place all these technologies that we think are needed for the next 10 years—not in one year or two years but the next 10 years.
How do you balance the consumer market with the industrial and medical markets? How do you get people to understand that designing for the long term is important?
If I didn’t have to deal with Wall Street, my ideal business would be an infrastructure kind of a business. I wouldn’t have to deal with cyclical ups and downs. Unfortunately, we have to deal with the reality of the business world and the fact that Wall Street is involved. We have to participate in fast-growth markets. And fast-growth markets by definition can also be fast-declining markets. When the GDP [gross domestic product] goes up by 1%, your business goes up 20%; it amplifies the GDP by that much. The infrastructure markets don’t amplify the GDP by that much. So, I love infrastructure businesses, but I don’t think we have a choice. We have to play in the mobile markets.
Does operating as a fab-lite company help with these big swings in demand?
Absolutely. I think that you cannot be in a mobile market having only your own fab. Consumer markets go up and down by so much, and they are very seasonal. How do you forecast your products if you’re completely in the consumer markets? Being a fab-lite company really addresses that difficulty. If you sell to stable infrastructure markets, you might keep a fab at 90% utilization. If your fabs aren’t full all the time, they won’t pay for themselves. Utilization factors have to be 90% plus, which is hard in the consumer markets. That is why fab-lite works so well for us.
As technology gets more complicated, does the job of semiconductor companies also get more complicated?
Yes. Our customers are paying a lot of attention not just to the chip that we provide, but also to the service that we provide. They seek the knowledge that we bring to the interaction with them.
| Reference |
|
