The article on where chips are headed was an excellent tutorial that helped bring me up to date a bit. Your question about IBM buying IDTI and manufacturing the C6+ to compete head-one with Intel brings up a possibility, but most likely a very small one. The inclusion of copper metal technology to increase processor speed would have to be answered by a real technologist. Many years ago, when I started with Bell Labs, there was something called copper poisoning that prevented processing engineers, of which I was one, from using copper. Aluminum, with its reduced electrical conductivity, was selected because it was a reasonable compromise that worked. I'd like to know just what IBM has patented, the line dimensions they can employ, and many other details of copper metal. Being skeptical of marketing
claims never hurts. For instance, can they make multi-layer metals with sufficiently narrow line widths using copper?
The other parameter of the C6+ as made by IBM is it is a 2.5V technology. How much of the associated logic and memory chips will work with 2.5V? TTL used to be a 5V system, and it has been modified to run using 3.3V. As transistor line widths move down to .25 microns, the channel oxide thicknesses also scale down, and they won't tolerate 3.3V. That's why a 2.5V supply has to be used, but all the components must work with 2.5V. The computer I'm using probably wouldn't run with 2.5V, and there are design problems in trying to mix voltages. The important conclusion of all this rambling is that the C6+ may not be a drop-in replacement for the C6. I'd like to know the answer to that question. Some things may be apples to oranges.
IDTI's C6 could gain acceptance in sub-thousand dollar computers, but it is not designed to be competitive with the larger die, more complex Pentium II microprocessor at the high end. Would IBM use the C6+ to enter only the low cost computer market? Perhaps. I don't care what you smoke, Fugazi. I just want to know if you inhale! |
