Is this bad memory?

Nebell

2[H]4U
Joined
Jul 20, 2015
Messages
2,379
My dad's old computer suddenly started restarting just as it is about to boot to Windows.

Everything else works fine so I thought maybe he has gotten a virus and I reinstalled Windows. Windows booted fine and I got inside and started downloading updates when suddenly it restarted halfway through Creator's Update.
I got into Windows again and continued updates but as I was updating nVidia it restarted again and suddenly started bootloop like before.

CPU temp seems to be about 30-35 degrees Celsius which is perfectly fine.
Mind me this is a very old computer, about 8-9 or so years old. It's still running an old Core i7 920 :)
I was thinking about replacing the whole thing but if this is ram issue then it would be cheaper just to replace that, everything else works fine.
 
My vote is power supply. Even a top tier brand would be flaking out by now.
 
There have been reports of memory issues on older X58 boards. I wouldn't rule out the memory (probably not the actual memory but the memory controller). Definitely worth running memtest for a while as it isn't hard to do.
 
Ok so I did the following.

1) Tested each ram stick individually (6x2gb).
2) Hooked the whole system on the psu of my computer.
3) Hooked the ssd to my second working computer.

1) Didn't help (can't be that ALL 6 sticks are faulty?)
2) No change.
3) The SSD works fine.

This rules out the SSD, memory and the PSU.
Which leaves me to:

1) CPU overheating.
2) Motherboard bugging (had northbridge overheating issues since I got this computer 9 years ago).
3) Video card (which is an old GTX 570 but I doubt it's faulty.

I vote the motherboard.
 
Leave all 6 sticks in and see if it passes memtest instead of individually.
 
Leave all 6 sticks in and see if it passes memtest instead of individually.

Well I could try that tomorrow. I'm in bed already (I'm in Europe).
Testing all this was pain in the ass. I had to disassemble my mining pc which has 6 video cards <.<

Hope you don't mind me asking, but what would be the point?
If it passes memtest, it just means that I'm correct in assuming it's probably a motherboard?
And if it doesn't pass, I have to try individual sticks which I already tried.
And if they all turn out to work (which they did with my latest check), it just means that it probably is the motherboard acting up ;)

And I've already ordered a new motherboard, cpu and ram.
It's a 9 year old computer. To me that's way past its time.
 
It could be the memory controller on the chip if it's been overclocked for a while. When you have six sticks in it will push the memory controller harder.
 
I'd check the thermal paste on the cpu as well.

I had a similar issue on a very, very old computer (p4 2.8/512/533), and was never able to figure it out even though it ran fine for over a year. Swapped all the components out to a twin of it and the twin ran fine. o_O And I took the original and configured it similarly using different parts and it works fine now too. Weird.
 
Not an identical case, but I have an i7 2600K that suddenly started giving memory errors. Turned out I had to give it a little bit more voltage to the IMC and it would work fine again. So on this the IMC is a bit worse for wear after some years running and needed a little pick-me-up. Maybe try something similar? A 6 memory stick load can be troublesome on a good day with X58 after all.
 
Back
Top