The press reports on this story were fairly well mangled. What IBM developed was
certainly not a chip that could run at 500Ghz... The Georgia Tech-IBM team demonstrated a
single transistor that was capable of switching at 350Ghz at room temperature and 500Ghz when cooled to near absolute zero (using liquid helium). This is a far cry from a chip running at 500Ghz. For example, the individual transistors which make up a Pentium 4 chip are capable of running much much faster on their own than the "clockspeed" the chip itself is able to maintain.
[H]ow fast an individual transistor can be switched and how fast you can clock a complex digital circuit that consists of millions of those transistors communicating in lock-step are two very, very different numbers.
Another very important distinction that didn't really make it into the major news media coverage was that this type of transistor isn't even designed for use in a digital circuits, but in Analog RF devices. The best immediate thing this promises is a walkie-talkie that can change channels very, very quickly (and even using the technology for this is a long, long way off).
That's all not to say it wasn't an impressive achievement, and that the techniques used can't be applied to future digital transistor design... But, there's some serious misinformation about this achievement out there right now.
Author
Topic: 500 Ghz chip, and potential for 1Thz (Read 2236 times)