In my experience, the processor score is always low after the initial installation. I think that it runs the processor test while it is still installing, so the processor is busy during the test.

Do not bother changing the Priority. For all practical purposes, it will have no effect. The reason is that Windows hides most of the functionality to avoid problems. You are doing a lot less than you think when you change the Priority, unless you use special utilities.

Lawrence

Ok then if changing the priority might cause problems then by all means do not do so. I am using an Intel P4 3.2 ghz processor and for me my scores and processing speed increased when I changed my priority to high.

This does bring up a point - Is windows tasking based on cooperative multitasking or preemptive multitasking? If it is cooperative then I can most certainly see why increasing a task’s priority would definitely cause problems most especially if just one poorly written service was running but I don’t see how this would necessarily behave the same way in a preemptive multitasking environment. At some point all services must be given some processing time to prevent task stravation.

It would seem to me that while HT will allow some mucking about with the priority without causing significant problems, priority changing should not have any real discernable effect on the total throughput of the machine.

I suppose that it is possible that changing the priority might give it a bit of an edge when running the benchmark process (thereby increasing the Processor Score) especially on a processor that has atypical background tasks. But this should be nominal.

Note that the WCG code is essentially Floating Point while most other normal non-gaming, non-graphical-processing programs are mostly scalar. This is an important consideration in that it allows better true processor utilization on HT systems when running the WCG tasks and other everyday non-floating point applications on the same system .

I understand that there is at least one person that has observed a greater processor score with the priority increased. I suspect that this is anomalous or that there is some other obscure cause. It does not, in my view, make the practice of changing the priority worthwhile or safe .

Beyond this, I have been wondering lately just how close the benchmark represents the actual working part of the process; in this case Rosetta. If indeed it does not adequately represent the process, then the CPU and derivative calculations (points) are all "off" . I know. That is a whole "nuther" can of worms and brings into question many other questions about the efficacy of the various processors relative to Rosetta.

I will not be surprised if a moderator yells at me for that one.

Regards All