Faulty module?

So, let me give you some backstory. I built my new rig about 2 months ago, always liked corsair's products so i got alot of corsair stuff:

Sapphire Radeon HD 7970 3GB GDDR5

ASUS P8Z68 Deluxe/GEN3

Intel® Core i7-2600K Processor

Corsair SSD Force Series 3, 240GB

Corsair Obsidian 650D Midi Tower Svart

Corsair HX 750W PSU

Corsair H100 Hydro Series

Corsair Vengeance DDR3 1600MHz 16GB CL9

Kit w/4x 4GB XMS3 modules, CL9-9-9-24, 1.5V,

Vengeance Heatspreader

From the get go the computer was running like a dream, no bluescreens no nothing. I used the XMP profiles from the start, and havent really looked back since day 1 of install.

However, today i was playing battlefield 3 - And all of a sudden i get the mother of all bluescreens and instant reboot.

Booted windows again, and once more it bluescreens after a while of playing BF3. After the second restart it kept bluescreening after welcomming screen of entering windows.

So i put the XMP profile off (Auto). And now the computer seems stable again. I ran memtest from a usb drive after reactivating the XMP profile. And got like 600+ errors in the first 10 minutes. Rebooted and disabled XMP and ran memtest again. Ran it for 20+ minutes without a fail.

What seems strange to me is that ive not changed anything in the setup or drivers (apart from maybe a couple of windows updates) yet somehow the XMP profiles are no longer doing their thing.

Any suggestions before i start flashing bios and testing the memories individually?

Any suggestions before i start flashing bios and testing the memories individually?

Nope, this should be your next step. But if memory is going to fail it will fail at any speed and not just one.

Flash your BIOS to the latest version and run your modules individually. May want to check your PSu while your at it. This could very well be a power issue too! :)

Some results then:

Flashed bios, tried running 4 modules in 1600mhz. Failed horribly in memtest from the start, windows not booting again.

Tried 1 module at the time. No fails, let it run for a few mins then tested the others. Also did 2 by 2 and tried all 4 slots on the motherboard. Stable as a rock.

So the only way i can stress these memtest fails is if i run all 4 modules over 1333mhz. On 1333 and lower they are working fine.

Currently running 2 modules (8GB) @ 1600mhz. Seems odd to me, can this be a power issue then? And how can i test if it is? (Other than buying a new PSU, which seems rather overkill considering i can get 2x better rams for the same money and still be a happy camper)

Sounds like you need more memory controller voltage. It's usually labled QPI/VTT or something similar depending on the BIOS since it only fails with all slots filled. More modules is more strain on the memory controller and may not be able to run on default voltages.

Some MB call it VCCSA and some have both . Check your MB manual to see which one it is. Then set it to 1.2 volts and see if it isn't stable then.

So update again, tried what u suggested peanutz94. Did'nt work. I thought id do more extensive testing with memtest86 from a usb bootable.

When i run all 4 modules @ 1600 mhz, i get fails instantly, about a minute into the test.

When i run all 4 modules @ 1333 mhz, it doesnt fail instantly. But around test 6-7 it starts failing.

The two modules i have in now (space 2&4 from the CPU), i made a full pass without a single fail @ 1600 mhz. Planning to do the same with the two i took out.

Leads me to belive that i have infact got 1 or 2 faulty module(s). I can't think of anything else, maybe the motherboard sockets are failing somehow?

Leads me to belive that i have infact got 1 or 2 faulty module(s). I can't think of anything else, maybe the motherboard sockets are failing somehow?

The fact that they all pass memtest leads me to believe there is nothing wrong with the memory. If it is going to fail it will fail at any speed. If you had a faulty module it would have failed in memtest when you tested them individually.

Seems like your issue is only when you have 4 modules in and could infact be a problem with the MB.

Ok, this is becomming quite an annoyance. I've been getting BSOD's and random resets with running only 2 modules now. Even at stock (1333mhz) speeds.

Starting to belive that its something else causing this issue, perhaps its related to the SSD?

Also tried a diffrent PSU (Fractal Design 800w) that i had in my old rig, managed to stress another bluescreen & random restarts.

As a final test i tried refit all the modules & change the SSD cable. Havent been able to stress the bsods or random resets. However, my computer will simply no longer boot with 4 modules.

Starting to belive that its something else causing this issue, perhaps its related to the SSD?

Thats very well possible. Maybe secure erase it and reinstall the OS and see if anything changes

But to me it sounds like your MB is the problem child. You may also want to pull the CPU and check the socket for bent pins. That sounds like a plausible possibility too!