Bruce wrote:
> Yes, it's a proof of concept, and the big news is that it runs at room
> temperature. But it was a full PowerPC implementation.
The article Glenn just pointed to claims it lacked an FPU and Memeory
Management support which would be needed to run a real OS on it, and
things like superscalar operation, branch prediction, and out-of-order
execution are also missing which would prevent it from achieving the
relative performance you would expect of a 1Ghz chip. As the article
points out, adding these things drastically increases chip complexity
and tends to drive clock rates down.
G.