Orthos Issues

enigmo

enigmo

enigmo

Associate
Joined
3 Jan 2007
Posts
462
Location
London
Am I right in thinking that Orthos running Blend mode on an unoverclocked E6600 and decent RAM (Corsair xms2-6400C4) shouldn't flake out on me after 12mins? And also, this mystical stress.txt file it insists I consult doesn't seem to exist either.
I've tried the CPU only stress to isolate the issue (20mins so far and still going, I know that's not long enough to be sure but it's still longer than 12mins) and I'll run the memory stress after, and maybe memtest86+ for a few hours after, but, off the bat, is this usual?
 
Check your ram and CPU voltages are set correctly as I know my E6600 defaults to 1.2V when loading optimised defaults in the bios but it should be 1.35V.

Also might be an idea to set your ram timings manually.
 
Not Orthos related but I I ran Prime95 (25.4v) and it failed 'blend' 4 times in a row after 6 hours exactly lol. Very strange to fail at the same time on 4 occasions. While Orthos runs fine 12 hours+.

Michael
 
Vcore seems to be at 1.325V. I've just restarted the Blend test as the RAM one seemed to be fine after 30mins or so. If it fails again, I'll push the CPU voltage up to 1.35V, check the RAM timings, and see if it's more stable. Thanks, and I'll report back :)
 
TBH the cpu voltage should be fine at 1.325V. Mine ran Orthos no probs 8hrs plus at 2.95Gig using 1.3V.

You could try running a few passes of Memtest86+ to check the ram on it's own.
 
Blend lasted for over an hour on my second attempt last night. I left it running when I went to bed and when I woke up this morning the PC had restarted :(
I'll try Small FFTs until it fails or for the next 12hrs, whichever comes first :)
 
Or until I'm TSing into my machne from home like a divot and it crashes out on small FFTs after 15mins or so and now I can't get to it until I get home... :p
Well. I might be able to say with reasonable certainty it's a memory thing, so tonight 12hrs of memtest might show something up. Is it possible before I start that there might be enough of a problem somewhere to knock over Orthos but pass memtest?
 
Doubt it, but computers do all sorts of weird things so who knows.:)

If you get no errors on memtest and it still fails orthos then I'd be thinking that it's either a dodgy CPU or the voltage to your north bridge is set too low but I still think it's a problem with your ram. Set the voltage on your ram to 2.2V just to be sure it's getting enough juice.

I take it all your temps are reasonable under load as too much heat can give you random errors?
 
Nearly 10hrs of memtest and no errors :) I'll try Orthos again with more volts for the RAM. Temps are all OK; 25Cish at idle, 45-50ish at load, never seen it go over 53C. If the extra RAM juice doesn't solve it I'll reseat the CPU and look at the northbridge. I'm just leaving the CPU until last cos these s775 coolers are a pain is the a$$ to remove and replace.
 
Right. I've given the RAM a little extra voltage (now at 2.2V), left CPU at stock (1.325V), and now Orthos has run for nearly 2hrs not falling over, so I must be along the right track (thanks Meths) :)
Oh, irony of ironings, I put BIOS P30 on it, the very BIOS that killed my server board and my SCSI card, and it works fine :)
 
You must log in or register to reply here.
Back
Top Bottom