Thorsten Froehlich wrote:
> You need double precision floating-point processing. Sorry!
I wanted answer almost identicall but I'm completly dumb in that area so did
quick search in google. It answered there are GPUs with 128-bit
floating-point precision.
ABX
