Alessandro Coppo <a.c### [at] iolit> wrote:
> For your information it is years that Java is just-in-time compiled and, as 
> I wrote on another post, benchmarks show that it can be a competitor to 
> Assemebler for numerical tasks.

  Apparently I am unable to run Java in JIT-compiled mode because whatever I
do, it's slow as hell. Apparently JRE 1.3.1 does not do this. I don't know
what else could I do than download the latest JRE.

  I once found a factorization program in the net made in Java, which
factorized arbitrarily large numbers using the elliptic curve method.
I gave it a really large number, which was composed of three very large
primes, and it factorized it in something like 2 minutes in my 1.2GHz Athlon.
I thought that "wow, that was fast!".
  Well, then I found the pari/gp library and utility, which is written in C
(with some parts optimized in assembler). It also supports arbitrarily sized
numbers and factorization using the elliptic curve method.
  I run it in a 500MHz UltraSparc, which is about half fast than my Athlon
(of not even slower than that). The program factorized the same number in
less than 10 seconds.
  I have been laughing at the Java program since then.

-- 
#macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb x]
[1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
-1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// - Warp -

Post a reply to this message