..."Our view is that there is a basic shift taking place in the semiconductor industry, whereby the driver for technological advancements is no longer the PC but other applications such as high-speed networking. This is a case in point,” he said. Intel is talking 20GHz microprocessors in 2007 that will only be possible with new forms of lithography and gate-oxide materials, the spokesman said, while IBM can take its bipolar HBT transistors with SiGe and yield 100GHz chips by 2003, without lithography shrinks."
I'm still investigating the question at this point. I do, however, want to echo the last statement of the article:
"“(Ask them) whether their announcement is based on concepts that have been previously published in the technical literature, or what makes it a breakthrough. (Ask them) about gate lengths and how they’re measuring ‘fastest’,” she said. "
What kind of chip are they going to make that runs a 100GHz? Will it be a logic microprocessor, or simply a networking/communication chip? Also, they were talking about transistor speed... which it was never indicated how they made the measurements...
All I'm saying is that they claimed a transistor speed, but offered little or no other data. I'm therefore reserving judgement until I see a LOT more information. |
