Le 05/06/2012 17:22, jhu nous fit lire :
> I compiled a static binary of Povray 3.7 RC5 for ARM because I just bought a
> Samsung Galaxy SII and was curious to see how fast the processor is compared to
> my other computers. I'll have benchmarks up soon. In the meantime I've posted
> the binary in the binary misc section so you guys can check it out.
> 
> 
For your numbers, with Intel compiler, on i7 @ 3.33GHz, --benchmark is

Parser Time
  Parse Time:       0 hours  0 minutes  0 seconds (0.543 seconds)
              using 1 thread(s) with 0.471 CPU-seconds total
  Bounding Time:    0 hours  0 minutes  0 seconds (0.000 seconds)
              using 1 thread(s) with 0.000 CPU-seconds total

Render Time:
  Photon Time:      0 hours  0 minutes  2 seconds (2.164 seconds)
              using 15 thread(s) with 2.486 CPU-seconds total
  Radiosity Time:   No radiosity
  Trace Time:       0 hours  2 minutes  8 seconds (128.999 seconds)
              using 12 thread(s) with 1525.728 CPU-seconds total

Notice that hyperthreading is little help, as restricting to 6 rendering
threads is not a doubling operation:

Distinct 6 cores:
Render Time:
  Photon Time:      0 hours  0 minutes  2 seconds (2.158 seconds)
              using 9 thread(s) with 2.483 CPU-seconds total
  Radiosity Time:   No radiosity
  Trace Time:       0 hours  2 minutes 38 seconds (158.196 seconds)
              using 6 thread(s) with 943.581 CPU-seconds total

Same 3 cores:
Render Time:
  Photon Time:      0 hours  0 minutes  2 seconds (2.160 seconds)
              using 9 thread(s) with 2.483 CPU-seconds total
  Radiosity Time:   No radiosity
  Trace Time:       0 hours  4 minutes 16 seconds (256.684 seconds)
              using 6 thread(s) with 1532.217 CPU-seconds total


(cores-control with taskset command)

Post a reply to this message