RE: near-term processing power growth

From: Eugene Leitl (eugene.leitl@lrz.uni-muenchen.de)
Date: Fri Feb 11 2000 - 18:17:33 MST


Ramez Naam writes:
> If TI really delivers this in 2010, it will be a 30x speed up in 10 years,

One should compare apples with apples. TI 'c6x MIPSen look good on
paper, but do not translate very well in real-time performance.

If you do MMX on a 1 kbit bus, that can be contstrued into a lot of
individual operations. That operations are not independant, and pretty
dumb at that, is of little relevant to press release mongers. Dead
tree can take a lot of numbers, without blushing.

> which would be 2 years earlier than expected. More realistically this
> project will slip (as all large scale computing projects do) by a year or
> two. Move the timeline out to 11 or 12 years, and this is right on the
> projected curve.
>
> Again, I fail to see strong evidence of deviation from an 18 month doubling
> time of computing power.

I don't see anything doubling but the amount of transistors you can
put on a die. This doesn't translate into anything too cool, as long
as you get special cases piled on top of each other (pipeline setup,
burst, keeping ALU/FPU filled with the right magic instruction
mix) for it to happen. Bah, humbug.



This archive was generated by hypermail 2b29 : Thu Jul 27 2000 - 14:03:39 MDT