Interview: Atiq Raza on clean tech, the Valley and AMD
Serial entrepreneur sees opportunities in batteries, processors
Rick Merritt
(05/14/2010 11:59 AM EDT)
URL: eetimes.com
SAN JOSE, Calif. — We sat down recently with Atiq Raza, entrepreneur in residence at Khosla Ventures, interim chief executive of battery maker Seeo (Berkeley, Calif.) and a veteran of Silicon Valley. The discussion ranged across topics including venture capital, the Intel-vs.-AMD battle, the outlook for clean tech.
EE Times: How would you characterize the state of the venture capital sector in Silicon Valley?
Atiq Raza: The VC model is undergoing a transformation. In the last ten years, a handful of VCs gave good returns in, another handful gave reasonable returns, but a very large number gave mediocre returns and an even larger number lost a lot of money.
So it's going to be very difficult for some VCs to raise funds. The total world of VCs is going to shrink dramatically. That’s not all bad. It will raise the quality of deals.
EET: Is innovation moving away from Silicon Valley?
Raza: There will be substantially less venture money going into Silicon Valley. That’s a good assumption to make. However I think innovation will increase in Silicon Valley over time. We will re-engineer ourselves into becoming experts in the new technologies.
EET: Like clean tech? What are you doing there?
Raza: I'm researching the possibility of helping to start a company that will address energy storage with a different scientific basis. There is a tremendous need in the world to make batteries have a higher energy density, be safe and have a high-speed way of charging and discharging.
We actually need two kinds of batteries. One kind is extremely high energy density--at least 10x what we have today--and we need a battery or ultracapacitor that allows fast charging and discharging. Classical batteries can't pull energy out for a sustained period rapidly because you are moving atoms and they can't move fast enough.
EET: How big is the battery opportunity?
Raza: It is the only industry in the world that will grow from $10 billion to $100 billion in the next ten years--and it can grow another five times that size in the following ten years. I don’t see an end in sight if there are sufficient breakthroughs, and even if the breakthroughs don't come, there will still be a hunger to absorb anything with an incremental edge.
EET: What markets are driving this?
Raza: Today wireless devices still drive the market for lithium ion batteries. There is about 200 million laptops worth of lithium ion battery capacity installed in the world today.
Tesla Motors uses about 2,000 laptops worth of batteries for one car, so 100,000 Teslas would consume the total lithium ion capacity in the world. So there will be a dramatic increase in lithium ion capacity.
As we try to supply cellphones for billions of people in the world and even several mobile devices per person its mind boggling what that will do the wireless infrastructure. The wireless tower will need battery backup. If you count all the energy requirements for all these applications [include smart grid backup] they actually exceed even the automotive requirements.
EET: Who are the likely winners?
Raza: From the best I can tell the Koreans are winning that race. I wouldn’t be surprised if in the next two years Samsung and LG emerge as the winners with Panasonic and Sanyo now merging and being left behind as number three.
EET: You used to be president of Advanced Micro Devices. What's your view of the Intel-vs.AMD battle these days?
Raza: The competition between Intel and AMD is being won by Intel right now. Intel is providing the leadership and setting the pace, but the industry and the economy would be better served if Intel and AMD were in a true footrace because they both strain harder to win. When Intel is so far ahead of AMD there is a tendency for neither of them to run that fast.
EET: What should AMD do?
Raza: AMD has to become a different company. It has to become a much more nimble, fast-moving company to close the gap with Intel. [AMD chief executive] Dirk Meyer has to give AMD the culture of a large startup rather than a small Intel. Dirk is a very creative person and can make AMD become that company.
Right now it is run as a smaller Intel. If it is a small Intel running against a big Intel, the big Intel will win.
EET: AMD has increasingly been paring back to just PC processors.
Raza: That’s the right thing for them. They have focused to where they have a benefit. But if they focus on only the lowest cost products, there will no longer be a race between Intel and AMD, just a partitioning of the market and of necessity in such a partitioning AMD will make less money than Intel.
The re-engineering of AMD has to occur, but it depends on how hard AMD wants to work. It's a much more comfortable path for AMD to just be a low-cost supplier and partition that part of the business and make less money but still make some. That will give them another 5-10 years of reasonable life, but I don’t know what happens after that for that kind of company.
EET: What do you think of AMD's split into processor and foundry companies?
Raza: That was an economic necessity. To fund the next generation of process technology and build a fab around it, you have to generate a certain amount of cash and volumes and revenues and AMD slowly was not able to meet the bar. It had to divest from its fab.
The future of foundries is their customers will be AMDs and TIs and Freescales. None of these companies in the long run have the ability to generate enough cash to live up to Jerry Sanders old motto that only real men have fabs. There are no real men left except maybe Intel and Samsung. Now there are hybrid men.
EET: Now that you have sold off RMI, do you foresee other opportunities for multicore processor startups?
Raza: Yes, it's possible to design better multicore processors. [The cellphone is] exploding in workloads and unit demand. If markets are growing, there is always an opportunity to do something better.
Whether I am able to come up with something better I don’t know, but it's intriguing me a lot. Starting another [multicore] company is a step further than I have thought, but I can imagine things that can be done that would make a big difference. Talk to me in say three years. |
