Hi,
I was very interested in that article, but disappointed
that the new approach on the NForce3:
http://www.tomshardware.com/motherboard/20040420/nforce3_250gb-03.html
was not benchmarked. But I was also disappointed that a LANParty benchmark was not run. I wonder how well THG is equipped to run benchmarks on multiple machines simultaneously. I would have put a high end graphics card in each of the gigabit-ethernet equipped machines, loaded Quake, hooked them all into a gigabit switch, and done a frames per second test to see if the performance and CPU usage made a difference. A complicated matter normalizing out the other effects, but THG is clever.
One more thing: In MidTown Madness, I notice that the performance increases noticably if I go into the Win2K task manager and set the priority of the task to "High". Do you guys ever do this when you do game benchmarks on hardware?
Thanks for Tom's Hardware! I don't make a purchase without it.
-Tom Schaefer
I was very interested in that article, but disappointed
that the new approach on the NForce3:
http://www.tomshardware.com/motherboard/20040420/nforce3_250gb-03.html
was not benchmarked. But I was also disappointed that a LANParty benchmark was not run. I wonder how well THG is equipped to run benchmarks on multiple machines simultaneously. I would have put a high end graphics card in each of the gigabit-ethernet equipped machines, loaded Quake, hooked them all into a gigabit switch, and done a frames per second test to see if the performance and CPU usage made a difference. A complicated matter normalizing out the other effects, but THG is clever.
One more thing: In MidTown Madness, I notice that the performance increases noticably if I go into the Win2K task manager and set the priority of the task to "High". Do you guys ever do this when you do game benchmarks on hardware?
Thanks for Tom's Hardware! I don't make a purchase without it.
-Tom Schaefer