Re: Xeon HT or not HT

2004-07-22 Thread David Purton
On Thu, Jul 22, 2004 at 01:33:50PM -0400, Dragan Cvetkovic wrote: > Greg Folkert <[EMAIL PROTECTED]> writes: > > > On Thu, 2004-07-22 at 05:10, nx13372 wrote: > >> I'm using kernel 2.4.26-1-686-smp. > >> I have a dual xeon box. If in the bios i enable the HT i'll get 4 cpus, > >> if not i'll get

Re: Xeon HT or not HT

2004-07-22 Thread Dragan Cvetkovic
Greg Folkert <[EMAIL PROTECTED]> writes: > On Thu, 2004-07-22 at 05:10, nx13372 wrote: >> I'm using kernel 2.4.26-1-686-smp. >> I have a dual xeon box. If in the bios i enable the HT i'll get 4 cpus, >> if not i'll get 2 cpus. >> What is bettter? [snip] > > Without hyperthreading you have 2 Thi

Re: Xeon HT or not HT

2004-07-22 Thread Derrick 'dman' Hudson
On Thu, Jul 22, 2004 at 10:10:28AM +0100, nx13372 wrote: | Hi all, | | I'm using kernel 2.4.26-1-686-smp. | I have a dual xeon box. If in the bios i enable the HT i'll get 4 cpus, | if not i'll get 2 cpus. You have 2 Physical CPUs regardless. With HT each physical CPU is divided into 2 Logical

Re: Xeon HT or not HT

2004-07-22 Thread Kirk Strauser
On Thursday 22 July 2004 12:18, Greg Folkert wrote: > I think the answer is clear. I'd have to disagree. We're running some Xeon servers with and without HT, and while there is *some* performance increase in some circumstances, it doesn't seem to be anything to write home about. I won't go so

Re: Xeon HT or not HT

2004-07-22 Thread Greg Folkert
On Thu, 2004-07-22 at 05:10, nx13372 wrote: > Hi all, > > I'm using kernel 2.4.26-1-686-smp. > I have a dual xeon box. If in the bios i enable the HT i'll get 4 cpus, > if not i'll get 2 cpus. > What is bettter? Think about your question. Speaking Hypothetically on idealogical designs: Without