"commodorejohn" <com### [at] gmailcom> wrote:

> > Actually, the N270 has hyperthreading, so 3.7 would show some gain with it.
> Ah, interesting. Just for the sake of completeness, here's the results of the
> 3.7 benchmark on the Eee:
>
> Parse Time: 0 hours  0 minutes  3 seconds (3.844 seconds)
>               using 1 thread(s) with 3.796 CPU-seconds total
> Photon Time: 0 hours  0 minutes 20 seconds (20.453 seconds)
>               using 4 thread(s) with 20.468 CPU-seconds total
> Trace Time: 2 hours 43 minutes  27 seconds (9807.313 seconds)
>               using 1 thread(s) with 9787.750 CPU-seconds total
> And a total time given of 9812.92 seconds.

Something's not right. Either povray isn't properly detecting the number of CPUs
or you're not running an SMP kernel. Here's what I get on my Acer Aspire One
(N270 @ 1.6 GHz):

povray --benchmark -WT2
Render Time:
  Photon Time:      0 hours  0 minutes 21 seconds (21.580 seconds)
              using 5 thread(s) with 25.332 CPU-seconds total
  Radiosity Time:   No radiosity
  Trace Time:       1 hours 43 minutes 53 seconds (6233.633 seconds)
              using 2 thread(s) with 12354.025 CPU-seconds total

povray --benchmark -WT1
Render Time:
  Photon Time:      0 hours  0 minutes 24 seconds (24.673 seconds)
              using 4 thread(s) with 24.655 CPU-seconds total
  Radiosity Time:   No radiosity
  Trace Time:       2 hours 52 minutes 17 seconds (10337.888 seconds)
              using 1 thread(s) with 10316.611 CPU-seconds total


There's a significant speedup with 2 threads vs. 1 thread.

Post a reply to this message