Message 86146 - Python tracker

Message86146

Author	pitrou
Recipients	alexandre.vassalotti, amaury.forgeotdarc, christian.heimes, eric.smith, gvanrossum, jaredgrubb, mark.dickinson, nascheme, noam, pitrou, preston, rhettinger, tim.peters
Date	2009-04-18.23:43:06
SpamBayes Score	5.3003147e-05
Marked as misclassified	No
Message-id	<1240098286.9193.6.camel@fsol>
In-reply-to	<1240097735.38.0.021023061845.issue1580@psf.upfronthosting.co.za>

Content
> Is there a way to use SSE when available and x86 when it's not. Probably, but I don't think there is any point doing so. The main benefit of SSE2 is to get higher performance on floating point intensive code, which no pure Python code could be regarded as (the CPU cycles elapsed between two FP instructions would dwarf the cost of executing the FP instructions themselves). The situation is different for specialized packages like numpy, but their decisions need not be the same as ours AFAIK.

Content

> Is there a way to use SSE when available and x86 when it's not.

Probably, but I don't think there is any point doing so. The main
benefit of SSE2 is to get higher performance on floating point intensive
code, which no pure Python code could be regarded as (the CPU cycles
elapsed between two FP instructions would dwarf the cost of executing
the FP instructions themselves).

The situation is different for specialized packages like numpy, but
their decisions need not be the same as ours AFAIK.

History
Date	User	Action	Args
2009-04-18 23:43:07	pitrou	set	recipients: + pitrou, gvanrossum, tim.peters, nascheme, rhettinger, amaury.forgeotdarc, mark.dickinson, eric.smith, christian.heimes, alexandre.vassalotti, noam, jaredgrubb, preston
2009-04-18 23:43:06	pitrou	link	issue1580 messages
2009-04-18 23:43:06	pitrou	create