On 15 Nov, 05:32, William Stein wrote:
> On Sat, Nov 14, 2009 at 9:00 PM, Sebastian Pancratz wrote:
>
> > Dear William,
>
> > I am adding a somewhat more detailed performance report below,
> > comparing my own FLINT-based C code, MAGMA, SAGE 4.1.2.alpha2 and SAGE
> > 4.1.2.alpha2 with the patc
When the CPU is in power saving mode it runs much slower. Usually they
are set up to run at full speed after a short burst of sustained use.
This can affect short benchmarks of course.
Bill.
On 15 Nov, 12:50, Gonzalo Tornaria wrote:
> On Sun, Nov 15, 2009 at 3:32 AM, William Stein wrote:
> >>
On Sun, Nov 15, 2009 at 3:32 AM, William Stein wrote:
>> model name : Dual-Core AMD Opteron(tm) Processor 8220
>> cpu MHz : 1000.000
>
> Why do you say "1000" MHZ when that particular processor is a 2800Mhz
> (=2.8Ghz) processor?
> [...]
>> model name : Intel(R) Core(TM
Dear William,
Thank your for the reply. I hadn't used PARI for this before, but
when I tried this right now I am getting very similar results.
About the strange CPU info, I actually cannot quite see where the
discrepancy is coming from. As perhaps you guessed from the format,
in each case I ha
On Sat, Nov 14, 2009 at 9:00 PM, Sebastian Pancratz wrote:
>
> Dear William,
>
> I am adding a somewhat more detailed performance report below,
> comparing my own FLINT-based C code, MAGMA, SAGE 4.1.2.alpha2 and SAGE
> 4.1.2.alpha2 with the patch from trac ticket #4000.
>
> To give an impression
Dear William,
I am adding a somewhat more detailed performance report below,
comparing my own FLINT-based C code, MAGMA, SAGE 4.1.2.alpha2 and SAGE
4.1.2.alpha2 with the patch from trac ticket #4000.
To give an impression of the results, for very small examples MAGMA
marginally beats my C implem
On Nov 14, 2009, at 5:59 PM, Sebastian Pancratz wrote:
> Dear all,
>
> Some might remember that in September I put a lot of effort into
> rewriting Q[t] to use FLINT. While the patch (see trac #4000) was
> very usable in practice already, despite the help of many people there
> remained a few do
On Nov 14, 9:05 pm, William Stein wrote:
> Could you add something to this email about performance comparisons with
> Magma?
> It's one thing to say "Sage is slow compared to my custom optimized C
> code" (no surprise), and another to say "Sage is 1000 times slower
> than some easy-to-write int
On Sat, Nov 14, 2009 at 5:59 PM, Sebastian Pancratz wrote:
>
> Dear all,
>
> Some might remember that in September I put a lot of effort into
> rewriting Q[t] to use FLINT. While the patch (see trac #4000) was
> very usable in practice already, despite the help of many people there
> remained a