QX9650, stock volts (I think) 750i FTW board H50 cooler
Everything - CPU-Z, Everest, OCCT, HWMonitor - reads vCore as 0.9V (stock), regardless of what I set it as in the BIOS. The BIOS does however read current voltage as what I've set it as.
I don't know what the voltage actually is though. I got to 3.33 fine on stock volts, when I figured out it wasn't changing, I set it as 1.3625V in the BIOS (max. safe), and ran P95 again, and this time I got a lock up after about a half hour, where-as it was prefectly stable before. Max. recorded temp was 51*C.
Any ideas? Aiming for 4.0 here.
Did you try anything in the 1.2 to 1.3 volt range before trying the max? Also have you checked to see if tweaking the RAM timings and voltage would fix it?
Yup, tried 1.25 and 1.3, and yup, adjusted RAM volts from 1.9-2.1, as well as loosening up the timings from 4-12 all the way to 6-18.
Has that MB OCed well for you before? If it's not the board or a driver problem, don't what to say. I assume you have the latest MB chipset driver and the CPU HS (or water block in your case) is seated firmly?
What's your motherboard? I have the exact same issue with a QX9650 on a DX48BT2 (BoneTrail 2) board. The BIOS reflects changes to Vcore, but CPU-Z, Everest and CoreTemp all report the factory VID (1.25V in my case).
I've never bothered to take the time to determine if this is a software reporting issue or, as I suspect, a hardware issue that Intel has never corrected, but I have a definite wall around 3.5GHz, give or take a few hundred MHz as a result.
The only way to really tell is to use a multimeter to measure the actual voltage being supplied. But this is way more effort than 4GHz is worth to me ;)
(Apologies for not finding information on where to measure Vcore at--I'm quite positive that it's possible, but it's just shy of midnight and I've been reading law all day. I'm going to bed now!)
Yes, I have all the latest drivers, yes, the H50's seated properly.
It's a 750i FTW from EVGA. I'm getting a wall around 3.5 too. I might try just sticking it at 4.0/1.4, see what happens...
Oh hell, I read that in your first post as a PSU for some reason. And if your chip is going to behave anything at all like mine, nothing will happen ;)
I don't know what the voltage actually is though. I got to 3.33 fine on stock volts, when I figured out it wasn't changing, I set it as 1.3625V in the BIOS (max. safe)
You can go over 1.3625v with good cooling like your H50. Remember, setting 1.3625 in the BIOS won't get you 1.36v actual because of vdroop. You may have to set it in the 1.37-1.38 range depending on your board. What is CPU-Z and/or HWMonitor reporting for your voltage under load?
What voltage are you using for the FSB/NB? On my EP45-UD3R i had to start raising the FSB/NB voltage once i got to around 475. You could also try bumping up the VTT and see if it helps, don't go to high with VTT though as it can kill your chip pretty fast. Try dropping the multi down and raising the FSB, that should be able to tell you if its your board or your CPU holding you back.
4Ghz out of QX9650 should be no problem at all. My Q9450 could do 4.0Ghz with 1.39v, and i even had it up to 4.2Ghz for a couple benching runs. I'm having similar problems with my i7 860 and P55-UD6 board, it hits a brick wall at 4.0Ghz. No matter what settings i use, or how much volts i give it, it refuses to boot at over 4.0.
Everything, barr the BIOS, gives me 0.9V, whether idle or under load.
I'll try 3.6/1.4 and see what that gets me.
Sounds like it could be reporting the mosfet voltages, or the NB voltage. I've had that happen on a couple of 775 boards before.
Might want to flash the bios to something else. You could have a board thats starting to die.