Velocity Reviews - Computer Hardware Reviews

Velocity Reviews > Newsgroups > Programming > Python > RE: make faster Richards benchmark

Reply
Thread Tools

RE: make faster Richards benchmark

 
 
Simon Wittber
Guest
Posts: n/a
 
      05-14-2004
>> And Simon: "I just tried the benchmark with Psyco, and cut the run
>> time for input=10000 from 8.1 seconds to 3.4"
>>

>What platform is this on?


WinXP, P4 2.6GHZ, 512MB RAM

Sw.


 
Reply With Quote
 
 
 
 
Peter Hansen
Guest
Posts: n/a
 
      05-14-2004
Simon Wittber wrote:

>>>And Simon: "I just tried the benchmark with Psyco, and cut the run
>>>time for input=10000 from 8.1 seconds to 3.4"
>>>

>>
>>What platform is this on?

>
> WinXP, P4 2.6GHZ, 512MB RAM


Hmm... did anyone else confirm your numbers, so that I'm
the odd man out?

I'm using Python 2.3.3 and the latest download of Psyco
for it.

If you're using that too, the only apparent difference is
the CPU. Is it possible that could make such a significant
difference? (Unlikely at best, I'd say.)

-Peter
 
Reply With Quote
 
 
 
 
Meno
Guest
Posts: n/a
 
      05-14-2004
Peter Hansen <(E-Mail Removed)> wrote in message news:<(E-Mail Removed)>...
> Simon Wittber wrote:
>
> >>>And Simon: "I just tried the benchmark with Psyco, and cut the run
> >>>time for input=10000 from 8.1 seconds to 3.4"
> >>>
> >>
> >>What platform is this on?

> >
> > WinXP, P4 2.6GHZ, 512MB RAM

>
> Hmm... did anyone else confirm your numbers, so that I'm
> the odd man out?
>
> I'm using Python 2.3.3 and the latest download of Psyco
> for it.
>
> If you're using that too, the only apparent difference is
> the CPU. Is it possible that could make such a significant
> difference? (Unlikely at best, I'd say.)
>
> -Peter



On AMD XP 2500 and 512 MB RAM: gone from 855430 ticks to 503353 ticks
with psyco. That is I think 41% less time.

Peter, note that getting it to run twice as fast "only" means 50% less
time. Maybe that is what you should compare with your 38%...

Nemo.
 
Reply With Quote
 
Peter Hansen
Guest
Posts: n/a
 
      05-14-2004
Meno wrote:

> On AMD XP 2500 and 512 MB RAM: gone from 855430 ticks to 503353 ticks
> with psyco. That is I think 41% less time.
>
> Peter, note that getting it to run twice as fast "only" means 50% less
> time. Maybe that is what you should compare with your 38%...


Thanks for the comparison. When I say 38% speedup, I mean a
reduction in time of 27%, or I mean it now runs at 73% of
the time it used to. We _are_ using the same meaning for
"speedup" here...

My raw numbers are 862806 ticks beforehand, which corresponds
reasonably well with your value. After inserting the two lines
"import psyco" and "psyco.full()" as the very first lines in
the file, I get 627476 ticks. Am I putting it in the wrong
place? (Also tried it at the end just before the call to
main() with no difference.)

-Peter
 
Reply With Quote
 
Duncan Lissett
Guest
Posts: n/a
 
      05-17-2004
Peter Hansen <(E-Mail Removed)> wrote in message news:<(E-Mail Removed)>...

> My raw numbers are 862806 ticks beforehand, which corresponds
> reasonably well with your value. After inserting the two lines
> "import psyco" and "psyco.full()" as the very first lines in
> the file, I get 627476 ticks. Am I putting it in the wrong
> place? (Also tried it at the end just before the call to
> main() with no difference.)


Pysco is now shown
http://www.lissett.com/ben/bench1.htm

http://www.lissett.com/ben/bench3.htm
 
Reply With Quote
 
 
 
Reply

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
gcc, spectral-norm benchmark faster without optimizations Melzzzzz C++ 2 10-04-2012 10:37 AM
Does this benchmark make sense? Farhad Farzaneh Ruby 3 05-31-2007 07:23 AM
make faster Richards benchmark Duncan Lissett Ruby 11 06-01-2004 08:03 PM
make faster Richards benchmark Duncan Lissett Python 15 05-14-2004 10:14 PM
Re: Richards bench benchmark Peter Hansen Python 10 05-06-2004 11:02 PM



Advertisments