Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: Elmer who wrote (169384)	8/13/2002 3:01:17 PM
From: Yousef	Read Replies (1) \| Respond to of 186894

Elmer, Re: "Do you think they will have problems yielding or just poor binsplits? Will SOI be the generic process for the foundry?" During the foundry "ramp-up" phase, AMD will experience overall yield problems. Longer term, AMD just won't be able to offer as high a "speed" processor for fear of bin split yields/stability. At .13um, SOI won't be a generic process offered by the foundry. Foundries are particularly driven by cost ... At .13um the foundry "generic process" doesn't even use Epi wafers. Make It So, Yousef

To: Elmer who wrote (169384)	8/13/2002 7:12:11 PM
From: Road Walker	Read Replies (2) \| Respond to of 186894

Semiconductors Stretching Silicon's Limits Arik Hesseldahl, 08.13.02, 6:15 PM ET NEW YORK - It used to be that Mark Wolf couldn't get anyone to listen to him. "No one had time," says the chief executive of Amberwave Systems. The "no one" he refers to includes dozens of semiconductor companies, who in years past were too busy trying to keep up with overheated demand for chips of all kinds. And what Wolf's company, New Hampshire-based Amberwave, was pitching to them was a technology known as "strained silicon," which can help make chips run faster and consume less power. More concerned with their immediate needs, the last thing chip companies could afford to do was allocate a team of engineers to figure how they could apply strained silicon to their own chips. But times change. The chip industry is limping ever so tentatively back to life, spending money on research and development along the way. Today the world's biggest semiconductor company, Intel (nasdaq: INTC - news - people ), announced that its next generation of PC microprocessors will be made with a handful of new manufacturing technologies, including strained silicon, and will debut in chips in the second half of 2003. In the strained-silicon process, the main material used to make computer chips is stretched slightly to let electrons move more freely. The result is that they can move faster, increasing a chip's performance, but also allow the chip to consume less power. Intel says introducing strained silicon to its manufacturing process will add only 2% to the cost of processing a wafer. With Intel embracing strained silicon, other chip companies are sure to follow--and are likely to come knocking on Amberwave's door. A few already have. United Microelectronics (nyse: UMC - news - people ), Taiwan's No. 2 contract chip foundry, has already said it is working with Amberwave to offer strained silicon to its customers. And with UMC using the technology, it's a fair bet that Intel's chief rival in the PC microprocessor business, Advanced Micro Devices (nyse: AMD - news - people ), will have access to it as well. AMD and UMC are jointly building a new chip factory, or "fab," in Singapore that will start pumping out chips in 2005. Taiwan Semiconductor Manufacturing (nyse: TSM - news - people ) also plans to offer strained silicon to its customers. Other companies working with strained-silicon methods include Hitachi (nyse: HIT - news - people ) and Toshiba in Japan, and IBM Microelectronics in the U.S. Strained silicon has its roots in research done by Gene Fitzgerald, first at Bell Labs, then at the Massachusetts Institute of Technology. Fitzgerald launched Amberwave in 1998 with backing from Adams Capital Management, Arch Venture Partners and Dow Chemical (nyse: DOW - news - people ), among others. Wolf says Amberwave is working with other companies he can't name to bring the technology to other kinds of chips. "We'll license our technology to anyone who's interested." The technique applies not only to PC microprocessors, but other kinds of chips, including digital signal processors used in mobile phones and other wireless devices. Strained silicon also integrates easily with other recent advances in chip manufacturing, Wolf says. Amberwave is hard at work on perfecting techniques for combining strained silicon with silicon-on-insulator techniques, which introduce a think layer of glass to keep electrons in the right place on the chip. The intention is to keep Moore's Law--which states that the number of transistors that can be fit on a chip, and therefore its performance, doubles roughly every year--alive. The laws of physics have always had a nasty habit of getting in the way of that progress. And like other companies that specialize in semiconductor intellectual property, relatively fixed costs and low overhead mean licensing revenue goes directly to the bottom line. Chip companies looking to have better chips when business improves--sometime next year, the conventional wisdom says--have a lot more time than they did before. "We've been very busy," Wolf says.