craigmnet@cow.physics.wisc.edu wrote:
>Hmm, surprisingly I found that my version was about 4 times faster on
>two different architectures.
And eventually, so did I once I revised my testing code. My mistake
entirely... I got running times of 2.537 vs. 0.5115 for a 1024x1024 array.
{ alpha OSF unix 5.3 Nov 11 1999}
Todd