I ran the futermark benchmarks. I ran 3dmark2001 - and i got like 14480 or something. I'm thinking - thats a decent score. I go and search and compare other projects... some people have the fsb at 255mhz! pentium 4 - with qdr thats over a 1000mhz FSB. The motherboard was like an older one. It wasn't canterwood. Then i ran 3dmark2003 - those scores seem normal - i got like 4710. Then I ran pcmark2001 - this one is the best out of all of them. i get decent scores. I got like 7600 on cpu, 6300 on memory, and 1400 on hard disk. I look it up i swear someone had like, N/A for cpu, 16000 for memory and like 18000 for hard drive. ? that makes no sense.
i have raid 0 - 2 pretty fast IDE drives which match up pretty good to a scsi drive. i get 1400 . a scsi drive i can see getting 1500, and in raid, maybe 4000 .. but i'm seeing scores above this including 18,000 and i'm tihnking how can that be? it makes no sense.
can someone explain these insane scores? are they hacked or what?
also my problem seems fixed - i just replaced an ide ribben cable which was hooked up to a zip drive and a dvd-rom. I also raised the mem voltage from 2.5 to 2.6. I did that cos most of the memory i see uses a voltage of 2.6. so i wonder if that was causing some problems too. It seems great now. So i guess "endpoint format is invalid" paired with "delay write failure" is a bad ribben cable and maybe even memory voltage too low.
Hey! Well done on finding a solution - I'd read about your problems in your various posts about 'em... :smile:
I was chasing a problem for 3 days last week - multiple BIOS flashes with 3 different BIOS versions, replacement disks, loads of other crap I went thru... and in the end, it was a combination of a dodgy ribbon cable and a dodgy molex power connector... Boy did I feel stupid. :redface:
But anyway, the HDD score is <i>possible</i> if they were using one of those PCI cards that use SDRAM to pretend to be a hard disk... But I usually take the cynical viewpoint, so I reckon they're probably faked somehow...