Loading ...
Sorry, an error occurred while loading the content.

11986Re: [PrimeNumbers] has SB been tested?

Expand Messages
  • mikeoakes2@aol.com
    Apr 1, 2003
      In a message dated 01/04/03 13:25:21 GMT Daylight Time,
      decio@... writes:

      > While I do agree that such testing would be reassuring and consume few
      > resources, it is my experience that coding an FFT for the 1 million digit
      > range is rather trivial.

      As is coding an FFT for 100 Million digits: size has nothing to do with it.

      What is /not/ trivial is to optimise it to the extent done for the GMP
      library incorporated (I believe) in the Proth search programs used by SB team.

      To maximize speed, George Woltman's FFT code incorporates an unprecedented
      amount of optimisation, and is constantly riding on the extreme edge of
      failing (roundoff >= 0.5 bits), so it needs an exceptional amount of testing,
      on many different architectures, especially near the many FFT-size
      See e.g. http://www.mersenne.org/source.htm

      Mike Oakes

      [Non-text portions of this message have been removed]
    • Show all 8 messages in this topic