speeding up + ctfe question

jerro a at a.com
Sun May 27 22:00:00 PDT 2012


On Sunday, 27 May 2012 at 17:00:01 UTC, maarten van damme wrote:
> ok, can't seem to reproduce the crashing. now on to optimizing 
> my
> sieve a bit more ,9 miliseconds for 1_000_000 is still to slow.
> --
> Er is zon buiten. Zonnige zondag namiddag met priemgetallen in 
> de
> plaats van buiten zitten. Tss tss. :-)
> hoe? wie? x)

Have you tried turning optimization on? I get about 8ms with
your code with no flags and about 2.9 with -O -inline -release.

To optimize it further you could make a sieve that only works
with odd numbers. I get (https://gist.github.com/2817289) about
1ms that way. I'm sure it is possible to do much better, but
that would probably be much more work too.


More information about the Digitalmars-d-learn mailing list