[an error occurred while processing this directive] [an error occurred while processing this directive][an error occurred while processing this directive] [an error occurred while processing this directive] [an error occurred while processing this directive] [an error occurred while processing this directive] (none) [an error occurred while processing this directive] [an error occurred while processing this directive] [an error occurred while processing this directive] [an error occurred while processing this directive] [an error occurred while processing this directive][an error occurred while processing this directive] [an error occurred while processing this directive][an error occurred while processing this directive] [an error occurred while processing this directive][an error occurred while processing this directive] [an error occurred while processing this directive] [an error occurred while processing this directive] [an error occurred while processing this directive] (none) [an error occurred while processing this directive] [an error occurred while processing this directive] [an error occurred while processing this directive][an error occurred while processing this directive]
 
[an error occurred while processing this directive] [an error occurred while processing this directive]
Skåne Sjælland Linux User Group - http://www.sslug.dk Home   Subscribe   Mail Archive   Forum   Calendar   Search
MhonArc Date: [Date Prev] [Date Index] [Date Next]   Thread: [Date Prev] [Thread Index] [Date Next]   MhonArc
 

Re: [CPROG] parallelisering



Sune Fischer wrote:
> 

snip - snip

> 
> > Kernen "booter" så at sige kun processorerne og
> > tillader et api lag som
> > mpi at snakke med et lag i kernen, der så spawner
> > opgaver ud til
> > processerne. Derfor har de forskellige kerne
> > versioner ikke den helt
> > store indflydelse mht. optimering af dit program.
> >
> > Parallel er lidt fancy, men det er tit at man bare
> > skal tænke sig om for
> > at optimere, f.eks. at
> > 4*2^4 + 4*3^2 er dårligt men at 4(2^4+3^2) er bedre.
> > Tæl selv
> > operationer efter.
> 
> Algoritmen og implementeringen er tæt på det optimale,
> i det 90 procent af den er Fourier transformationer
> copy/pasted fra numerical recipies.

Hehe...det er nu ikke altid at den kære recipies er super optimeret, jeg
kan bare aldrig huske urlen til updates...er der nogen der kender den ?

> 
> > Endelig kommer vi til gode kompiler flag. Generelt
> > skal >-02 ikke bruges
> > på intel chips da gcc er kendt for at kunne lave
> > mærkelig coredumps over
> > denne optimering.
> 
> Jeg bruger kun -O3, men loop unrolling og alle de
> andre ser ikke ud til at spille en rolle, ikke et
> sekund vindes, de fleste kører bare langsommere.

Nemlig!

> 
> > -mpentium er fint og kører (vist nok) meget stabilt.
> 
> den må jeg vist lige prøve, irriterende den ikke
> automatisk kan optimere for MMX og andre gode
> instruktionssæt. Jeg går ud fra den stadig kompilerer
> til 386'ere? Det må altså kunne gøres bedre.

Den virkelig hardcore optimering ligger i at kalde registrene selv, det
kan de fleste kompilere ikke gøre for een.

> 
> Hilsen Sune
> 
> __________________________________________________
> Do You Yahoo!?
> Talk to your friends online with Yahoo! Messenger.
> http://im.yahoo.com

-- 
Venlig hilsen / Best Regards

--------------------------------
Bernino Lind
iSolve ApS
          sslug@sslug
--------------------------------
Rømersgade 7 kld
DK-1362 Copenhagen K
Denmark
[+45]  33 32 59 78
www.iSolve.dk
--------------------------------


 
Home   Subscribe   Mail Archive   Index   Calendar   Search

 
 
Questions about the web-pages to <www_admin>. Last modified 2005-08-10, 20:08 CEST [an error occurred while processing this directive]
This page is maintained by [an error occurred while processing this directive]MHonArc [an error occurred while processing this directive] # [an error occurred while processing this directive] *