doesn't help. I tried 1.60 and 1.65 for voltage too, but still same crash pattern. edit: btw, is it of any use to me that i run it multiplier 12 instead if 11? still shows up as 2500.. :\ edit2: how about updating bios? any chance that will help?
No, but what it will tell you is if its stable at its current speed. You need a base to wrok from, a stable one. If the cpu isnt stable at say 2ghz, then trying for higher is pointless as you dont have a point to work from.
followed your advice. Ran it for 5 hours or so with no problems to report. Do we go anywhere from here?
Was that at 166*12? on 1.6x v's? If so try something like 170 or 175. Might need to increase vdimm, so give it say 2.7v if you can. Run memtest for a good 30 mins to check its stable, then boot back to windows and prime again.
x12 Edit: ok, raise ram to 175.. How long do i have to run prime afterwards? 5 hours after each 5 mhz raise is cumbersome and btw, do ram sticks overheat? How would i know if they did? edit#6 - multi x11 and fsb for ram and cpu @ 175 ran fine with memtest for half hour and ok in prime for an hour. I'll try raise it to 185. edit #7 - @185 ran memtest for half an hour and prime crashed after 3 hours. However it didn't error out, just the windows error. I will run it again for some 5 hours tonight, and then depending on the results might try 190 tomorrow.
Sounds like you got things covered. Shouldnt overheat unless you are putting loads of volts through it
Or so i wish... Anyhow, prime seemed to fail again this time with a "FATAL ERROR - rounding paka paka". Not sure what this can be - this is @ 185 after about an hour of operation. i have ram at 2.7 v and vcore 1.675. The only thing (based on my rather limited knowledge) i could adjust is the voltage, but looking at the histogram i saw no profound voltage drops when the test failed. I guess tomorrow morning i'll try some final changes and see if it passes. If not, whatever, back to 2500 There goes my career Vlad edit: i think instead of messing with the voltage i'll run a longer memtest, make sure its not the memory thats at fault. Wish me luck, good people.
One way to get an indication of whats failed in prime is to look at the fft leght that the test failed on.... larger (ie about ~ 256-300k) are likely to indicate chipset or mem issues, under that, could indicate cpu But yes, try memtest for a longer period. It can be infuriating sometimes, trying to get these 2 rigs stable atm, and its having none of it lol!
6 hours of memtest, not a single error. I think the ram is stable, which explains why i got errors when the multiplier was 12.5. either way, any suggestions i should explore further? edit 1: 1.70 voltage, still prime errors out, and cpu heats to some 64 degrees according to asus probe. I changed some settings in bios and ganna try this one last time tomorrow. edit 2: Hmmm funny. I played with some settings, and removed one case wall. First dropped temp by some 9 degrees, and second went ok with prime for 3 and a half hours. My plan is to rasie to 195 and run the memtest and prime 5 hours each. updates shall come.
Doh! Doh! I'm under the unpleasent impression i killed my ram. @195 memtest ran fine, so i decided to up to 200. there it ran fine for an hour. I decided to see if i could boot to windows. First time i got an odd error about some file missing, second time black screen, and third time... Well, 1 long beep and no image on screen. Checked to see if it was the vid card, but it ain't. I'm guessing its the ram, but i dont have spare sticks to check. So we'll see tomorrow...
Clear the CMOS, its most likely the board wont boot at 200mhz into windows. Can sometimes be the case, memtest stable, but no boot to windows...
You ruined it Highland3r. I was really looking forward to buying a new ram stick Well this was fun. So what's your opinion, i should look into buying a 400mhz ram stick? Thanks, Vlad
