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

A question Benchmarking primes.

Expand Messages
  • padmapani19
    I am wondering if there exists a benchmark with which i can compare two prime-generating algorithms.for that matter even for factorisation algorithms.i assume
    Message 1 of 2 , Apr 22, 2002
      I am wondering if there exists a benchmark with which i can compare
      two prime-generating algorithms.for that matter even for factorisation
      algorithms.i assume the answer would be a function of space and time
      complexity.but i am ignorant of anything like that.

      This came up because in my class i have to compare my algorithm with
      that of somebody's.i want to know how people in real world stack up
      their rankings of algorithms.

      Padmapani
      *-The sufficiency of my merit is to know that :
      my merit is NOT sufficient-*
    • Phil Carmody
      ... Minimise the number of variables. Make sure that the things that are not important for the comparison are the same. i.e. if it s the high level algorithm
      Message 2 of 2 , Apr 22, 2002
        --- padmapani19 <padmapani19@...> wrote:
        > I am wondering if there exists a benchmark with which i can compare
        > two prime-generating algorithms.for that matter even for
        > factorisation
        > algorithms.i assume the answer would be a function of space and
        > time
        > complexity.but i am ignorant of anything like that.
        >
        > This came up because in my class i have to compare my algorithm
        > with
        > that of somebody's.i want to know how people in real world stack up
        > their rankings of algorithms.

        Minimise the number of variables. Make sure that the things that are
        not important for the comparison are the same. i.e. if it's the high
        level algorithm that is to be compared, then use identical low-level
        libraries. Standardise on GMP or LIP, or Miracl. Unless part of the
        assignment is "write fast low level primitives", of course. If it's
        the implementation as a whole that's being measured then chose or
        write your low leve llibraries carefully, and the only thing that
        matters then is the clock.

        When I'm benchmarking, I often put a counter to each function and
        count how many times each primitive is called, and that's enough to
        satisfy my curiosity. Of course, I am ignoring issues such as the
        ones Yves mentioned, those of caching etc., but my main interest is
        in the ~2^64 arena presently.

        Phil

        __________________________________________________
        Do You Yahoo!?
        Yahoo! Games - play chess, backgammon, pool and more
        http://games.yahoo.com/
      Your message has been successfully submitted and would be delivered to recipients shortly.