| 1 |
On 11/8/06, Alan McKinnon <alan@××××××××××××××××.za> wrote: |
| 2 |
<SNIP> |
| 3 |
> > It's pretty boring but it seems that you can sort of double the CPU |
| 4 |
> > MHz spec and come pretty close to the BogoMIPS numbers. However that |
| 5 |
> > doesn't take cache size into account so maybe BogoMIPS isn't even the |
| 6 |
> > right thing to be looking at. |
| 7 |
> |
| 8 |
> bogomips is usually about double the cpu speed, but you can't count on |
| 9 |
> that. I would imagine that cache size and fpu performance were |
| 10 |
> significant factors. |
| 11 |
> |
| 12 |
> Let's assume this app of your is floating point intensive (fairly safe |
| 13 |
> assumption), and doesn't use a heck of a lot of RAM or disk (already |
| 14 |
> shown to be true). So now you need to rate the fpu of the various |
| 15 |
> processors and machines around. So I would read reviews of various |
| 16 |
> machines in computer performance mags where they publish sane |
| 17 |
> benchmarks, to get an idea of what would be best |
| 18 |
> |
| 19 |
> alan |
| 20 |
|
| 21 |
Yeah - makes sense. Thanks. |
| 22 |
|
| 23 |
Back to Google FPU stuff. |
| 24 |
|
| 25 |
Thanks for your inputs. |
| 26 |
|
| 27 |
Cheers, |
| 28 |
Mark |
| 29 |
-- |
| 30 |
gentoo-user@g.o mailing list |
Archives