My nForce HD Benchmarks

[CraigJay]

My nForce is up and running with an Athlon XP 1900+. I have a 40gig Seagate Baracuda IV as the C: drive and an 80gig Seagate Baracuda IV as the D: drive (both as masters on different channels). There were no other drives in the system during these tests. For the test file I used a 674 meg CD image.

Copy from Motherboard IDE1 to Motherboard IDE2: 18 seconds
Copy from Promise Ultra100 IDE1 to Promise Ultra100 IDE2: 48 seconds
Copy from SIIG Ultra100 IDE1 to SIIG Ultra100 IDE2: 50 seconds
Copy from Motherboard IDE1 to SIIG IDE1: 37 seconds
Copy from SIIG IDE1 to Motherboard IDE1: 20 seconds
Copy from Motherboard IDE1 to Promise Ultra100 IDE1: 36 seconds
Copy from Promise Ultra100 IDE1 to Motherboard IDE1: 19 seconds

I draw 2 conclusion: The motherboard IDE controllers are fast on the nForce. Bothe the SIIG and the Promise Ultra100 controlers read fast, but write slow.

I'd be happy to run other benchmarks on my nForce for anyone interested. Please provide a link where I can download the software to run the benchmark if that is needed.
Running Windows XP Pro
MSI nForce Board
2x256meg Crucial CAS2
Athlon XP 1900+
onboard sound
onboard video
onboard LAN (Copy files between another computer that is slower at 7.7meg per second)
Seagate Baracuda IV 40gig as C:
Seagate Baracuda IV 80gig as D:
Pioneer DVD 116 as E:
Plextor 24x writer as F:

The only software problem I have found is with BattleCom and Game Voice. They simple dont work in some games, and on some games they cutout very bad. This is with games that worked just fine on my old PC. I suspect immature drivers, as I've been able to screw around and get some games to work OK.

Craig

 

[trikster2]

Craig

Thanks for all the work!

Wish more reviews would do simple real world tests like this.

Could the slow performance of the PCI cards indicate a slow PCI bus on the NForce similar to the recent VIA problems? (vs slow cards).

Also: what guarantee do you have that each itteration of the test was using the same cylinders? Comparing read/writes from inside tracks vs read/writes on outside tracks would give the hardware a fair chance.

Again thanks for taking the time to run some benchmarks we can all easily relate to!

 

[CraigJay]

Before copying the file, I would do a fresh reboot. After copying the file, I would delete the copied file. The file is always in the same source spot, because I never moved it. Nothing was ever changed on the second drive, where I copied the file to, because after I copied the file to that drive, I deleted it, emptied the recycle bin, and rebooted for the next test.

Craig