Hi! I mainly run my processor at stock since I'm not 100% sure of its stability under OC... and I haven't yet had a chance to run Prime95 for the >24 hours that has been recommended to me - though its run overnight for around 10-14hours several times with no problems.
I just ran the "Intel Burn Test v2.4" on 'high' and it ran for only several minutes (getting my core0 temperature to 69C, while prime95 only ever got it up to 61C). When it finished it said it failed!
Is this a reliable overclocking test? ...I just found a link and recommendation for it somewhere. What should I try now. I have my i5-750 at 190x20=3800Mhz and the vcore at 1.325volts. I am wary of increasing it further. Should I continue upping the vcore? Are there other settings I should try on my GA-P55-UD3R?
If you click the about button in the program it will state how it can cause your cpu to run up to 22% hotter than prime 95. Those are real errors with your o/c.
Do you maybe have power saving states on, and or turbo ?
The program stresses memory highly as well, what is your memory speed timings?
I need more voltage at 3800 with my 750. If you don't want to add more vcore, scale back your bcclck to 3700 and then 3600 if necessary. Voltage increase does not rise linear. I noticed a higher increase at 3800 and even more at 4000 and if I want to run at 4200 it would be close to 1.47, so I don't. http://www.tomshardware.com/reviews/core-i5-750-overclo...
My memory is at 10x - so it's running at 1900Mhz which is very slightly over spec for it (4GB of F3-15000CL9D-4GBRH).
@zipzoomflyhigh: is it possible for it to fail the test simply because of the temperature? Or are you suggesting that the temperature would cause problems in the cpu that would cause it to fail the test?
I was told that <75C is a safe temperature for the i5... not that I was happy to see it hit 69C (it was mostly at 65-67, it spiked once).
Any other suggestions, etc? Is the consensus that I should not raise the vcore and instead give up on the 3.8 OC for a 3.7 or 3.6 that might work at the same vcore?
Alright. I did as advised. I cranked the tried both 3.6Ghz and 3.7Ghz by changing the baseclock to 180Mhz then 185Mhz. I also tried first reducing my memory multiplyer from 10x to 8x, which would put it well under spec, to see if it was a problem with the memory - it was not.
Now 3700Mhz is pretty good, so I'm not disappointed, but I'm wondering if there would be any way to get it stable at a higher clock speed (...maybe lowering the cpu multiplier...)? Perhaps there are other settings I could try: this is my first time OCing and I have only made significant changes to the vcore and baseclock. Perhaps another voltage is the culprit.
I don't want to increase the vCore much more, if at all, and I don't want to risk the longevity of the cpu too much (perhaps 3700Mhz is the best option for a longterm OC). My cooling situation is essentially a "Cooler Master 212 Hyper Plus" mounted with "Rosewill RCX-TC060 Thermal Paste", and 3 additional case fans in my antec 300 (the 2 std. aiming out, +1 extra aiming in). Does my temperature seem high for this CPU? (Keep in mind that prime95 only ever gets it up to 59-61C and it idles at 32-34C, at 3800Mhz) I was careful, but could I have mounted it poorly?
Your temps are fine. Max core temp is actually 98c and tcase(top of cpu) is 72.7
Most o/c's at 4ghz will touch 70 in prime 95, some better some worse.
Thanks. I had seen that video before putting the computer together, but didn't have the experience to pick up all the details. That was interestesting to see their temperatures: 80C, wow! He didn't seem worried. I'm guessing they were just trying to get it OCed quickly so they went overboard with the vcore and that is why it is so hot. They were using liquid-cooling too. Isn't that supposed to keep it much colder?