On Tue, 09 Aug 2005 15:41:35 GMT, "Donald McTrevor"
<me@privacy.net> wrote:


>Maybe they did, maybe Tom's hardware got it wrong, either way
>I will found out soon when I replace my Cyrix MII 300 which
>unfortunately does not feature in Toms tests.

There is no "maybe" too it. These are old technologies so
we have the luxury of hindsight. Many of use have (or still
do) own a few systems with their processors and can be
certain of how much slower the K6-2 is at everything, but
again this is only considering (every parameter of) CPU
performance, not performance of a motherboard chipset or
video card or (other bottleneck).

>
>Also I believe CPU's can me made to 'target' certain benchmarks.

Because they have different architectures, indeed any given
CPU will be faster at some task than others. However, some
things remain constant. For example, the lack of L2 cache
on a K6-2 makes it always slower than K6-3 at same MHz
speed, same FSB. K6-3 is never as fast as a P3 either, not
in any benchmark that leaves the CPU as a significant
bottleneck. There is nothing that I'm aware of that any
K6-2 can do even 70% as fast as the P3 in the linked
article.

Benchmarks are only as good as the subsystems they're
stressing. You wouldn't use a benchmark for monitor
response time to gauge CPU peformance either.