[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: G5 bandwidth

Subject: Re: G5 bandwidth
From: Drew McCormack <email@hidden>
Date: Fri, 23 Jan 2004 09:00:18 +0100

Small memory code (110 MB): Time Opteron / Time G5 = 2.3
Opteron: 85 seconds
G5: 36 seconds

Large memory code (450MB)  Time Opteron / Time G5 = 1.9
Opteron: 454 seconds
G5: 237 seconds

While the G5 is faster, the Opteron seems to have a better memory
access/bandwidth.

Could it simply be that the reason the opteron gains on the g5 in the second test is simply that the G5 is a much more efficient chip (for this problem), and you are making it wait more in the second run, due to the bandwidth constraints? It doesn't matter how fast your chip is, if the bottleneck is bandwidth, because it will just spend its time waiting. Just a theory.

Drew McCormack
_______________________________________________
scitech mailing list | email@hidden
Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/scitech
Do not post admin requests to the list. They will be ignored.

Prev by Date: Re: making dynamic libraries with gcc
Next by Date: Re: G5 bandwidth
Previous by thread: RE: G5 bandwidth
Next by thread: Re: G5 bandwidth
Index(es):
- Date
- Thread

Home