|
From: | Marcus G. Daniels |
Subject: | Re: Ok, what about benchmarks |
Date: | Wed, 29 Jan 2003 14:13:56 -0700 |
User-agent: | Mozilla/5.0 (X11; U; SunOS sun4u; en-US; rv:1.3b) Gecko/20030117 |
There are some superficial and `easy' things to fix that can make big performance differences. It isn't fair to say some system is slow just because it runs a particular software package slower than another system. For example, one might discover that there was some diagnostic flag set (unintentionally) on one system that wasn't present on another. Another possibility is that one system implicitly had extra work (say, that Cygwin was doing CR/NL -> NL conversions). And sometimes when there is a narrow system weakness, it's avoidable. For example, the problem Paul mentioned with slow printfs on Windows can be fixed by running the shell in Emacs.I only did this side by side comparison when the guy I was working with asked how long a simulation run took for me and I told him a few minutes and he said that it was taking him hours.
When there is a factor of ten difference for similar hardware, there is probably just some mistake that can be avoided with a little investigation.
================================== Swarm-Support is for discussion of the technical details of the day to day usage of Swarm. For list administration needs (esp. [un]subscribing), please send a message to <address@hidden> with "help" in the body of the message.
[Prev in Thread] | Current Thread | [Next in Thread] |