Motherboard interaction with ram issue

[reavilak]

I have this ram motherboard issue that's killing me. Like around 2 years ego I bought 64 GB of ram for my Desktop. I work with WMwares and needed this ram for testing networks setups. Right after I bought the ram I started to get blue screams and my CPU started to reboot around 2times a day. At first I assume that the ram I bought was bad, so I used Windows mdsched to test it, but this tool found nothing wrong with ram. So I assumed there was nothing worng with it, not realizing at that time that Windows mdsched is a peace BS. Just FYI, at that time I had also installed several things into my Desktop, so I wasn't sure if the rebooting was a result of the ram, or the other things I had upgrade - like my HD and even a bus deference between my processor and the RAM sticks. My processor back them had like a bus of 1600mhz - or around that area - and the ram I bought had a bus of 1866. I had installed a CPU before in this fashion, with a ram greater than the CPU speed, and the CPU was stable. So back them I assumed the main culprit here was the interaction of the processor with the RAM for that specific CPU architecture. So, I decided to wait and buy a processor that matched the speed of the new RAM. Well, I did that, but my computer still reboots, with that ram( even though now its stable for almost 4 days before crushing ). Well, I decided to test the ram now using a more advance tool this - mentest86 - and this tool found a problem in the ram while doing the hammer test ( test 7). This error didn't appear with Windows BS tool mdsched. What is kill me about this problems, is that, I am still not sure about the problems nature. When I test the 8 slots of 64GB of ram at the same time, memtest86 detects a problem with the ram, when I test the ram sticks one by one, memtest86 tell me all the ram sticks are good :/ . I started to think it could be simply that the my motherboard was not that compatible with that specific ram. But now I have come to think that can't be the case( not a ram incompatibility with MB issue but maybe one ram stick is really bad ) because I have now tested all the ram stick in pairs, and when I do this, only one pair always starts to fail - even if the slot changed. now I am thinking, but not sure if this true, because I don't know enough about ram architecture, that there has to be a component inside the RAN sticks that helps them to communicate with the other RAM sticks, while connected and that this hypnoticall component is the one that is failing in one of my ram sticks. I have already detected the ram stick that is possibly the bad RAM. I have tested this ram stick in pairs, with other RAMs, and when tested in pairs both sticks fails, but individually it always passes. I am so confused, I am about to through away that ram stick, but I am not sure, I just don't want to through a good ram stick, without really being sure it's truly bad, but with this problem I just can't tell for sure.

My CPU configuration is : 64gb corsair vengeance 8gb sticks with 1866bus; motherboard x79-ud5 and i7-4930k

 

[cin19]

Recommend to use XMP profile for your RAM, you can set it up in the BIOS memory section. Then update the chipset driver, motherboard BIOS too. Keep in mind, when you update the BIOS, don't use 64GB RAM, only use one stick (8GB), if the PC reboot during the BIOS update, that will brick the MB.

Also what GPU and PSU did you have? because PC reboot problem will be caused by PSU, driver, overheat, etc too?

 

[reavilak]

I am using the 13w bios for the x79-ud5. I have already updated the bios to the latest version. Just FYI, the 13w bios is a beta version bios, but it's the only bios that works with the chip that I have now, the i7-4930k. Moreover, I don't think the problem is the BIOS, even though this its a beta version, because before buying the 4930k chip I used to have the i7-3820 processor, with that chip I used to have the BIOS version 10, which was not a beta BIOS, and the problem was still there( the problem was even worse then; it rebooted like 2 times per day, after the change, now only like once every 4 days). Like I pointed out in my initial post, that chip had a max of ddr3 of 1600( the3820). That is the chip that I changed with the 4930k because it had a speed that was the same to the ram( a speed of ddr31866) The point is that I don't think the bios is the problem. I have updated this several times with different chips and the reboot problem still persisted. Just FYI, I have a Nvida Gforce for the grafic card, but I don't think the problem is related to the card.

 

[cin19]

I have updated this several times with different chips and the reboot problem still persisted.

I mean the chipset driver https://downloadcenter.intel.com/download/20775 not the chips, whatever you mean 4930k or 3820, or 4930K.

And how about to use the XMP for the RAM too, because those ( bios update, driver update, etc) will help to solve the compatibility problem too.

 

[reavilak]

I have updated this several times with different chips and the reboot problem still persisted.

I mean the chipset driver https://downloadcenter.intel.com/download/20775 not the chips, whatever you mean 4930k or 3820, or 4930K.

And how about to use the XMP for the RAM too, because those ( bios update, driver update, etc) will help to solve the compatibility problem too.

Thanks for the help. I did update the chipset, but I have not changed XMP, I will try that too

 

[reavilak]

I have updated this several times with different chips and the reboot problem still persisted.

I mean the chipset driver https://downloadcenter.intel.com/download/20775 not the chips, whatever you mean 4930k or 3820, or 4930K.

And how about to use the XMP for the RAM too, because those ( bios update, driver update, etc) will help to solve the compatibility problem too.

Hello, I didn't realize this until I went back into my BIOS to check the XMP, but what u meant by XMP is simply overclocking the ram to its correct speed, right of 1866. Well, I have already done this, I did this when I realized the motherboard had timed my ram wrongly at around 1300 ( don't know why it didn't detect this automatically, that's bad ), when I sow this issue I activated profile1 ( which I think is what you meant by XMP; this PF had the correct speed of 1866, or actually if I remember correctly, it said 1867mhz witch is wrong, I corrected that to 1866, but welI, I traied with both and it failed). Well, when I did this the problem initially become worse, my CPU only rebooted until the Windows logon, them rebooted again. Then I tried to do the OC manually, without the profile thing, when I did that my CPU got more stable, and lasted longer without rebooting, but still rebooted. My desktop is most stable when I load optimized setting, it lasts like 4 days without rebooting.

 

[cin19]

My desktop is most stable when I load optimized setting, it lasts like 4 days without rebooting.

That means you need to tweak the cpu voltage if you oc the cpu. You may find the review for more info. Something like: http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/63119-intel-i7-4930k-i7-4820k-ivy-bridge-e-review.html

Also I don't know you bought those RAM is the whole kit ( 8 X 8GB) or two kits, or more ( like 2 of 4 X 8GB, or 4 of 2X8GB). If you don't buy the RAM in the whole kit ( 8 X8GB), that means they don't play well each other even under the XMP setting, so you just use the PC under the load optimized setting, or try to set the Voltage or timing but you will need time to tweak.