On Wednesday 26 November 2008 23:00:29 mabshoff wrote:
> On Nov 26, 2:51 pm, "Bill Hart" <[EMAIL PROTECTED]> wrote:
> > You are right. I think it won't be in GMP 4.3.
>
> Then the speed improvements for gmpbench are all related to the
> improved Opteron code? I have heard that allegedly there will be a lot
> more Opteron assembly code in gmp 4.3 which could contribute to the
> better scores.
>

Consider the redc fn in the file mpz/powm.c , this is just a mpn_mul_basecase 
with a "twist" , so we would expect much the same speed improvement that we 
get from going from a c to asm version of mul_basecase , and gmpbench uses 
rsa which uses mpz_powm. I expect gmp-4.3 has a K8 asm mpn_redc_basecase.


> > But there is nothing stopping us putting it into eMPIRe.
> >
> :)
> :
> > We can test against the FLINT FFT to make sure it gives correct
> > results. That has already been done at various points as Paul and I
> > were in contact about the respective FFT's, but we can do it again for
> > ourselves.
>
> Yep.
>
> > Bill.
>
> Cheers,
>
> Michael
> 


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"mpir-devel" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at 
http://groups.google.com/group/mpir-devel?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to