Performance issue?!

[Jordan K. Hubbard]

On Dec 23, 2010, at 8:14 AM, Marko Käning wrote:

> OK, I have given a lot of data comparing my MacOSX against the virtual Linux - i.e. using the same hardware.

Aha, yes, let's use the same hardware in both cases!

> Now, my 1year old iMac has a 3 GHz Dual Core.

Sure, OK.

> Let's compare this with my 1.5 years old cheap ASUS laptop. It is Dual Core T4200 at 2.2 GHz, i.e. definitely slower than my iMac.

Um, not OK? :)

If you're really serious about figuring out what's going on here, install Linux on your 1 year old iMac (not running under a VM, obviously, but in another partition) with BootCamp and actually compare performance on identical workloads, then compile the code you're benchmarking with profiling enabled (this is supported on both MacOSX and Linux) and compare the bottlenecks you see.  That would be quite instructive.  Comparing hardware by age or price is not, since all hardware has individual strengths and weaknesses (as do the OSes which run on it) and simply looking at CPU clock speed or memory bandwidth does not tell the whole story by any stretch of the imagination.  That is why you have to take all variability out of the equation before comparing X and Y for pretty much any value of X and Y or any conclusions you reach won't be worth the trouble you took to get there.

- Jordan

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.macosforge.org/pipermail/macports-dev/attachments/20101223/aac80303/attachment.html>