[ODE] Quick, quickstep comparision

Megan Fox shalinor at circustent.us
Tue May 18 16:53:04 MST 2004


Impressive - so at least now ODE is a far sight better than Newton.


That said, can anyone say why Tokamak performs consistently better?  Is it
their algorithms being superior, or is this just a case of quickstep still
being a little unoptimized and generally new code?

-Megan Fox

> stepfast1      quickstep   tokamak      newton
> 0 : 2.157ms    0 : 1.081   0 : 6.179ms     0 : 0.780ms
> 1 : 2.205ms    1 : 1.057   1 : 0.559ms     1 : 0.573ms
> 2 : 2.102ms    2 : 1.040   2 : 0.546ms     2 : 0.579ms
> 3 : 7.009ms    3 : 2.214   3 : 1.053ms     3 : 1.346ms
> 4 : 12.294ms   4 : 3.640   4 : 1.166ms     4 : 1.274ms
> 5 : 16.723ms   5 : 4.480   5 : 2.227ms     5 : 2.357ms
> ...
> 39 : 33.572ms  39 : 8.402  39 : 6.140ms 39 : 10.937ms
> 40 : 32.728ms  40 : 8.357  40 : 5.983ms 40 : 10.766ms
> 41 : 32.874ms  41 : 8.435  41 : 5.858ms 41 : 10.915ms
> 42 : 33.026ms  42 : 8.533  42 : 5.940ms 42 : 10.953ms
> 43 : 35.253ms  43 : 10.423 43 : 6.158ms 43 : 12.926ms
> 44 : 32.704ms  44 : 8.081  44 : 6.057ms 44 : 10.080ms
> 45 : 33.069ms  45 : 8.069  45 : 5.904ms 45 : 11.072ms
> 46 : 32.587ms  46 : 8.575  46 : 5.977ms 46 : 10.711ms
> 47 : 32.540ms  47 : 8.362  47 : 6.047ms 47 : 11.127ms
> 48 : 32.495ms  48 : 8.069  48 : 6.204ms 48 : 11.456ms
> 49 : 32.142ms  49 : 7.978  49 : 5.944ms 49 : 10.467ms
> 50 : 32.215ms  50 : 8.089  50 : 5.733ms 50 : 11.116ms



More information about the ODE mailing list