In article <tsfcv0dfu0aubbig7ttd8q72tmr9v3519f@4ax.com>, f/f george
<george@yourplace.com> writes

On Tue, 25 Jan 2005 05:29:19 GMT, no one <noone@yahoo.com> wrote:

In general, how much slower is WU results than when just one is run?

My son has HT and he slowed about 25% from the single unit speed.
He went from 2 hours for one unit at a time to 2 1/2 hours per unit
while doing 2 at a time.

This is my experience as well. 

Whether dual CPU systems, P4 HT systems or dual Xeons with HT.

On dual Ht activated systems there is a double whammy, due to the shared
FSB and RAM. Dual CPUs slow down by @ 20% - 25%. When HT is activated
there is a further slow down of about 50% (based on the single WI time.
However, the total output, per 24 hours, is still higher.

My experience on a dual 2.8 Ghz Xeon, reported before, is -

Single WU on 1 CPU at ~ 2.5 hrs each (9.6 WUs per day)

Two WUs on 2 CPUs at ~ 3.5 hrs each (13.7 WUs per day)

Four WUs on 2 CPUs + 2 HT CPUs at ~ 5.5 hrs each (17.5 WUs per day)
-- 
John Clark