Re: [Haskell-cafe] Re: Code and Perf. Data for Prime Finders (was: Genuine Eratosthenes sieve)

Hi

But for the current version of my code, there is still a bit of a performance gap between our two methods. Here are the stats I get (ghc -O3, 2.4GHz x86):

Are you aware that -O3 is slower than -O2 and -O in ghc? If you want "fast code" then specify -O2, not -O3. Thanks Neil