Re: filling big array of double

> > I'm testing on a P4 3Ghz and this is my code:
>
> Uh oh, things get scary: I'm getting 1672 ms with Pascal version
> and 1132 ms with my or Aleksandr's version on the AXP 2200+ !
>
> That would make the P4 rather slow... (even for a P4?).
> Are you sure there isn't something wrong with your P4, like automatic
> clock throttling because of overheating? Or maybe it's a P4 Celeron
> with very little L1 cache?

I'm impressed with the performance of my Pentium M 1.6 GHz as well:

ASM version: 1423 ms
Pascal version: 2053 ms

Twice as fast as a Pentium 4 with twice the clock speed.
.

Relevant Pages

  • Re: Lies, damn lies and benchmarks
    ... The only way to get a 386-16 to be twice as fast as a 286-16 was to ... compared to the previous generation processor, ... faster in every way to the 386 without requiring a clock ramp. ... Pentium was succeeded by the Pentium Pro/II/III family. ...
    (comp.security.misc)
  • Re: Core2 Duo Vs Pentium 4 2.6 GHz
    ... GHz model AND that the Pentium 4 is a 2.6GHz model, then obviously the dual core chip can do twice as much work as the Pentium 4 CPU; but it will only run twice as fast when you are running an application written to run dual threads. ...
    (comp.sys.intel)
  • Re: Benchmark results for Dothan
    ... > My first result is the Move benchmark (Pentium M Dothan 1.6 GHz): ... seems too good to be true (over twice as fast as every other function), ...
    (borland.public.delphi.language.basm)
  • Re: apt yast ...?
    ... >>an aging 1.2 GHz ... Only that it is over twice as fast as my Pentium II 350 whick runs fine ...
    (alt.os.linux.suse)