i7 2600K in a Maximus IV GeneZ/Gen 3

[Mgutierrez33]

I am currently having no end of problems with this rig. Until recently I was running a super shitty Antec Kueler h2o 620 as a CPU cooler, but still managed to have the processor clocked to 4.6 Ghz (auto settings in Asus BIOS), and temps never climbed above 70 C. When I transitioned from a Lian Li mATX case to a Corsair Graphite 600T, something changed during the component swap and now my temperatures went to shit. Every single time I would test it on Prime95, it would become unstable, blue screen,t hen restart. I messed around with the heatsink, reseated it, checked to see if it was making any unusual noises (it DID make something of a loud noise from the pump for a few minutes but then cleared up), etc etc. Did a Prime95 test on Large FFT's, every time it crashes once temps climb above 70 C, which can now happen. moved clock speeds back to 4.5 Ghz at 1.38V, everything seemed to be working better. Now I got a Corsair H100i in the mail the other day, installed that. Tried testing it today at 4.6 Ghz, 1.38V, temps are skyrocketting. Checked to make sure the pump was even running (HATE CorsairLink software, btw). Uninstalled the pump and noticed that one side did not seat properly, going to reseat and re-test again when I get the time (hopefully within the next few days.

Does 1.38 V for my overclock seem a bit high? Still learning the nuances of overclocking and want to make sure that I'm doing this all correctly as my experience is somewhat limited at present. Any other advice on how Sandy Bridge overclocks typically tend to go and what to expect and look for?

16-01-13

Well, to update, I have since reseated the heat sink and backed off my overclock entirely to stock. I am noticing that it takes my motherboard an exceptionally long amount of time to POST now and sits at the splash screen to enter BIOS for a long time. It does eventually boot, but when I get in and check Core Temp I see that my idle temps are 32 C on core #3, and as high as 39 C on core #0. Under Prime95 temps almost immediately climb to over 70 C across all cores. What the hell does this even mean, has my processor taken a shit on me? My temps are continuing to climb even as I sit here typing this, and I am doing nothing else in the background. If anyone has any advice on this PLEASE help me out :-(

 

[Lauralarry]

take your chip out inspect around the socket, I remember once I got thermal paste in the socket I was getting 97c on all my cores and immediately took my pc apart and seen that I added too much thermal-paste and it was everywhere

also uploading pics of your setup would help too, along with a snapshot of CPU-z next to Coretemp

things that could be wrong:

Broken Cpu Fan header
You're plugging the pump into a molex that's not proving enough power to it
You added to much Thermal paste
You didn't add thermal paste (lol)
There's thermal paste in your socket (Be glad you have Intel if this happened)
both pumps are broken (unlikily)
or your boards sensors are bad

 

[Mgutierrez33]

sadly I am currently having to type from a different computer (my backup in the event some faggotry like this happens), so screenshots aren't a possibility since I don't really want to run the machine more than is needed. Pulling up CPU-Z and CoreTemp reveals that at idle my rig is currently sitting at an average of 36 C, and under small FFT's (Prime95) I spike to temps as high as 81 C, all at OEM clock speeds and voltages (approx. 1.25 V). I have checked the pump in all of the fan headers on my motherboard by holding the pump of the H100i in my hand while turning on the rig momentarily: I can feel it noticeably torque in my hand, telling me that it is, in fact, turning on.

Since I have also now checked the pump on all fan headers I can also confirm that the fan headers are all functioning correctly (ASUS AI suite also detects RPM from all headers). ASUS temp sensor has always been a bit wonky, and never seems to mesh up correctly with CoreTemp or Real Temp GT (both have been used during testing to determine temperature discrepancies between programs, both read within 1 C of each other most all the time). I have also once again uninstalled the pump from the CPU to triple check the processor, socket, heat sink, and thermal paste. Paste is applied neatly and evenly across the processor, no gunk has gotten into the CPU socket, and no pins are bent.

Thank you, by the way, for the troubleshooting tips. It often helps to have a second outside mind approach the problem you are faced with to remind you of the things you know but forget x-).

From what I can determine/observe, it's looking like one of the following:
1. the thermal compound on the inside of the cpu itself has finally taken a shit because of the poor cooling solution I had to begin with, causing processor failure (it DOES run fully stable at stock clock speeds, but I dare not overclock it any more for fear of permanently putting the comp out of commission.)

2. Something is horribly wrong with the H100i I ordered. Harder to diag that one since the whole thing is a fully enclosed unit and can't reliably be disassembled to examine and test further. Even worse, having purchased this thing from Newegg it'll take approx. 798348603746980 years for me to RMA it if that IS the case -_-.

Given that this is a problem that has been degrading and getting noticeably worse over time, I am presently leaning towards the processor dying on me (this actually reminds me of the recent issue that Tom Logan reported having with his old test bench i7 he was using for heat sink testing. Gradually was incapable of holding any good known O.C. settings, and even got to the point that even at base clock speeds it was a bit wonky.). I was hoping it wouldn't get to that point as I got this 2600K when it was still commanding full original price and REALLY didn't want to have to replace it with a 3570K (all I would be able to afford, and only just... college student, so I'm broke). I would rather have waited until Haswell came out and gone to that. I'm rapidly running out of ideas though, and those are the only two parts that make any sense to me to have anything wrong with them :-\

Any thoughts?

Update: I have tested the H100i in a known good AMD rig which contains an 8320 O.C'd to 4.2 Ghz in a Gigabyte UD3 990FX board. With the fans hooked into molex and large FFT test run the temps never climbed over 43 C with the H100i installed (a full 8 C cooler than with the Seidon 120M that normally resides in the system). This leaves me with the processor or the motherboard being boned -_-