Need help to find fault in MSI 760GM-E51?

I have a couple of MSI 760GM CMV4GX3M1A1333C9 a-E51. Unfortunately, the system has recently become unstable, and I'm trying to isolate the reason for this. I am currently keeping track of my tests on the MSI forum, but recently I found myself in a discussion with a member of more than Memtest and my RAM. Could you clarify the situation of instability and memory?

The statement is making is that my RAM I cannot possibly be the reason why my machine locks sometimes, because if there was any failure of the participation of RAM would live constant BSODs. This does not match my own experience system instability with other systems, and looks like a reasonable generalization.

I have Memtest, and found errors, suggesting a problem with memory, certainly no sense to try an exercise in detection of failures in other components by using a piece of hardware (RAM), which is returning values wrong?

To clarify, I am completely conscious that the cause for failing Memtest might be due to the CPU or the motherboard, although I would like to make clear whether I am right in thinking that this memory problem is very important before you begin try something else.

On the one hand, a bit wrong somewhere can fly all to hell. But on the other hand, if that part is a part of memory that is not being used by some parts of the OS, then the program memory that can correct the error around, or just give corrupt data or something.
As I am not sure how they were thoroughly with memtest, you should try one stick at a time in every slot with Memtest86 +.

That's exactly what I thought, despite the "close to correct the error that" may well be an application operating with errors or throw exceptions, not exactly a BSOD, and do not necessarily show any visible trouble either. Thank you for clarifying, for a moment I thought I was going crazy.

It was an awkward mix of configuration memory, I have since tested the two groups individually, a set (the pirate) memtest86 + no, so I'm re-bats individual events at the moment. I realized it makes me look like an idiot, not knowing what I was thinking at that time.

The complete installation failed after about 14 hours, since then I've tried all the sticks individually and establish a Corsair stick with error only within ~ 2.5 hours, I’m flushing memtest to observe if the address is reliable or not incase this helps.
Configuration is in stock, everything set to auto and running in 1333. take a picture of the screen once you have a couple of mistakes and transcribe the details here. It has been suggested to attempt to block 1066, what do you think?

I think complete blocks are almost always related to memory. Especially if there are errors in memtest. Just my 0.02.

Ok, small trouble. I have two photos of memtest establish an error in this suit, although on both occasions memtest failed the test only once out of countless passes. Since the primary attempt, stick with errors, with the following

Quote:

---

Code
test 6, pass 2, failing address 000d1eba184 - 3358.7MB, good ffdfffff, bad: ffffffff, err-bits 00200000, count 1

---

and when I left the test had completed 8 passes in less than 9 am, with only a single error. I tried to limit the scope of the report on this direction and the restriction to test 6, but could not get it wrong.
I then decided to restart, change the memory IGP (32mb having up / down), and return to the suite of tests and just leave it much longer to see if I could get a consistent success in that direction. From the second attempt, stick with errors, with the following:

Quote:

---

Code
test 6, pass 14, failing address 000d1eba184 - 3358.7MB, good ffdfffff, bad: ffffffff, err-bits 00200000, count 1

---

Attempted 1066, memtest yet, this time management, test, and the error bits changed:

Quote:

---

Code
test 8, pass 7, failing address 00092397304 - 2339.5MB, good 55cd05fb, bad: 55c505fb, err-bits 00080000, count 1

---

Would the Command Rate to 2t and set the memory voltage to 1.7 volts and see if that solves the problem.

Tried it, left memtest and just make sure the results. 63hr Wall time goes to 56, three errors

Quote:

---

Code
test 8, pass 36, failing address 0010a2bc3a4 - 4258.7MB, good 17413fc1, bad: 17433fc1, err-bits 00020000, count 1
test 8, pass 50, failing address 000a48bcf0c - 2632.7MB, good b0ffeefd, bad: b0f7eefd, err-bits 00080000, count 2
test ?, pass 56, failing address 000d1eba184 - 3358.7MB, good ffdfffff, bad: ffffffff, err-bits 00200000, count 3

---

I do not really like the way his report an error beyond the 4 GB installed in the system. Updated with another error that occurred while I was writing my post, just before the end of the test. Unfortunately my picture cut the number of test, but I'm sure the pattern.

These errors suggest that memory is not me, and you must set the command rate at 2T manually and if you have a setting for the memory controller could try to increase the + 0.2 volt