5 May
2010
5 May
'10
1:02 p.m.
On 5/4/2010 2:06 PM, Tony Duell wrote:
When what you're comparing is orders of magnitude faster. You can argue
No, not even then. I've seen RISC CPUs that execute an intruction in one
clock cycle (and of course there are superscalar CPUs that do mpre than
one instruction in a clock cycle) and CISC CPUs which do essentially the
same instruction in perhaps 50 cycles.
They
weren't that slow, 27Mhz compared to the PS2's 33Mhz. Polygon-wise, the
Since when has the master clock frequency been a useful measure of
procesor speed? that CISC != RISC or that one CPU is totally different
than another, but
when I say your 16MHz processor is slower than my 2933MHz processor,
it's a useful measure.
You know, I think I could concoct a classic 1GHz Z80. All I do is take
this 1GHz clock and divide it by 250, (giving 4MHz), and use that to
clock a plain old Z80A. Well, ther'es a 1GHz clock in there, and it is a
Z80, and...
I sitll stand my my statement that the master clock rate is meaningless.
-tony