Paging file on seperate hdd causes errors?

[erasmus372]

System in sig, 2 different brand SATA hdds in no Raid array. Upon installing windows I always create a 3GB page file at the start of the second drive, and remove any on the system drive. My prime95 blend errors almost instantly, superpi32m fails within 2 iterations, sp2004 fails within 30 seconds.

However, I've done 33 full passes on my ram with memtest86+, prime95 small fft doesn't seem to fail, and superpi 1m-8m seems fine. I haven't had a chance to try it with the paging file back on the system disk, but is this a common error? How hard is it to find and use a paging file on a disk that has nothing else happening on it?

[CGR]

I always put my page file on a second drive. Never had a problem. Not sure why your pagefile would even cause prime to fail.

[El<(')>Maxi]

It could be bad hardware on the second disk, try running some diagnostics on that drive as well as placing the page file on primary drive.

[erasmus372]

Hehe oh man I bet you're right! I nabbed this second disk as a 'broken' hdd while repairing a coworkers tower and putting a brand new one in there. Rewrote the drive with all 0's, partitioned, and all the SMART events it had been getting suddenly disappeared. Hasn't given me any cause to think it is still suspect, but now that you bring it up!

[erasmus372]

Well tested that second hdd with some thorough tools from Seagate and it passed just fine.

Moved the page file to system drive and suddenly superpi32m, orthos, and prime95 no longer fail!

[Mr.Guvernment]

i dont see why the pagefile would affect CPU stress testing?

[El<(')>Maxi]

The page file & system cache are often working in relation to one another. I think the drive has a problem even though it passed the Seagate test suite.

[erasmus372]

i dont see why the pagefile would affect CPU stress testing?

It won't for small FFT's and lower pi's, but blend and 32M make use of your page file.