http://www.linuxhardware.org/article...51&mode=thread

Note that there is HT versus non-HT benchmarks, where the benefit is shown in the kernel compile. Other benchmarks show an ever-so-slight decrease in performance.

This is to be expected since the two logical processors share the same memory structures, reducing their effective size. The upshot is that there is a benefit in aggregate instruction throughput. Even if throughput does not result in direct performance improvement (one processor could simply be executing an idle thread) this then has the effect of reducing latency between each thread timeslice.

This results increased responsiveness and 'smoothness', even if it doesn't result in a 'performance' benefit. Most people that have had the luxury of running multiprocessor systems should understand, except that a HT or CMP processor brings this without the system expense.

I know that this has been a while since I've posted. Suffice it to say that I've been busy and have just been reading the forums and not posting.

My opinion only... as always.