IBM, Intel pace each other with improved transistors
IBM and Intel accelerated their horse race in semiconductors when each unveiled at midnight Saturday Eastern Time similar chip-manufacturing advances.
The research from both companies involves a crucial building block — called high-k material — to build smaller, more efficient transistors in microprocessors. High-k materials are better insulators than standard silicon dioxide, allowing engineers to keep shrinking transistors without losing efficiency through leaking electricity.
In both announcements, engineers say they plan to use the material to build transistors that switch on and off better, using “high-k metal gate” technology.
The announcements promise to keep alive Moore’s Law, which holds that the number of transistors on a chip doubles every two years. It’s good news for users because the more transistors that can be packed on a microprocessor, the faster it runs a PC.
The new materials also mean the manufacturers won’t have a problem continuing to etch transistors on chips at microscopic sizes, and more importantly, mass-produce them so they’re affordable to PC users. In fact, Intel officials predict this breakthrough alone will ensure Moore’s Law thrives “well into the next decade.”
The announcements underscore an old industry rivalry, since on its part, IBM worked with Sony, Toshiba and Advanced Micro Devices (AMD), Intel’s main opponent in the microprocessor market.
Both Intel and AMD say they will use the technology to speed the transition from 65-nanometer chip-building architecture to 45-nm. Intel is due to start in the second half of the year production of the family: a new laptop dual-core microprocessor, a desktop dual-core and a quad-core, and server dual and quad-core processors. It expects to ramp up to full production in three factories by the first half of 2008.
“That would only be the first five of 15 products that we have in development for 45nm,” said Kaizad Mistry, manager of the 45-nm program at Intel.
Pressed on when the company would actually begin selling Penryn chips, Intel declined to be more specific.
AMD plans to produce its first 45-nm chips in mid-2008, in the wake of the launch of its first 65-nm product, the quad-core “Barcelona” chip due out in mid-2007, according to AMD spokeswoman Jessica Kaiser. IBM said it plans to sell systems with chips that use the new transistors by the end of 2008, using its manufacturing line in East Fishkill, New York.
Intel insists it has a large lead on all competitors in the progress of shrinking chip features to 45nm.
“We expect followers, but no company is anywhere near where we are with this incredible advancement,” said Intel spokeswoman Kari Aakre. “We are not only announcing an amazing transistor breakthrough with our high-k, metal gate solution, but we’re also already demonstrating it with five early versions of processors from our Penryn family, including server, desktop and laptop systems booting all major operating systems and various applications.”
The new processors will mark a big improvement in speed and performance, and require less electricity, than the previous generation of Intel microprocessors. If all goes as planned, they will also give Intel a bit of a head start in mass-producing chips with 45-nm transistors. A nanometer is microscopic measurement, just a billionth of a meter. More than 2,000 transistors 45nm wide could fit on the period at the end of this sentence.
Intel’s manufacturing lead should allow it to bring chips to market before its rivals, but IBM could get a greater return from this technology in the long term because it uses the high-k metal gates in a different way, one analyst said.
“It’s a wonderfully parallel development of a technology that should lead to faster, more efficient chips in everything from PCs to cell phones and iPods,” said Richard Doherty, senior analyst with The Envisioneering Group.
“Intel has the advantage that they’re already in production, but IBM’s advance may be even more significant and lead to faster, smaller chips. The IBM breakthrough is to integrate the metal gate so it’s embedded in the silicon. Intel put the metal gates on top of a proven silicon architecture.”
IBM confirmed that it planned to use high-k, metal gate technology not to produce faster chips in the short term, but to solve long-range problems such as progressing from 45-nm production to 32-nm and 22-nm. The company will use the advance to aid development of large-scale servers and supercomputers, said Bernie Meyerson, chief technologist for IBM’s Systems & Technology Group.
“We don’t build Vespa scooters, we build Ferraris. We’ve been talking about high-k for five years now, and if we wanted to, we could ship it out the door tomorrow. But there’s no reason to do that because it doesn’t solve any problem for us. We’re not addressing a crisis issue that hit us in the head when we didn’t see it coming,” Meyerson said.
“Ours is a more fundamental implementation; it’s a drop-in, or a one-for-one replacement, for SiO2,” he said, referring to existing silicon dioxide technology. “I’ve said for years that gate oxide scaling is ending. The gates are literally five atoms thick. What are you going to do, build one that’s two-and-a-half atoms thick?”