Can RAM be bad even if it passes memory diagnostics?

[lightingft]

Hey guys,
I've been having random lockups for a while, usually once per 3 days, sometimes once a day. At first I thought it's my failing HDD. I removed it. No lockups for 12 days. Until now. It randomly locked up one time, then another after 2 hrs. The only thing I suspect now would be my RAM probably... I've had NMI Parity Errors randomly before..
Another thing is that Windows updated my PC, and it crashes at the same day, could that be the case? The only updates are Security updates tho.
Also RAM passed all MemTests... So the question is could it still be bad after passing the MemTests? RAM sticks are about 5 years old.

 

[kanewolf]

When you say "passed all MemTests" do you mean that you ran MemTest86 for several clean passes?

 

[skit75]

Which memtests?

You need a boot-up memtest for more reliability. The windows one works "ok" I suppose but I like memtest86's better. Make a boot disc and test with theirs.

 

[Shaun o]

If you are getting Nmi parity errors.

You need to check your memory sticks visually.
look on the sticker or label of each memory module you have.

If it states the memory is Ecc, then you need to enable the Ecc option in your bios of the motherboard if it exists.

If the memory does not state it is Ecc enabled on the sticker you need to again go into the bios and disable the Ecc option in the bios of the motherboard.

Nmi is an extra parity bit error, so it suggests that the memory you have is Ecc capable.
And the option in your bios is not currently enabled in the bios.

Ecc enabled memory does not like it if the motherboard bios does not support the feature for the memory to enable the extra parity bit it uses. And will throw up the Nmi error from time to time within windows.

If you find it to be the case then you will have to change your memory to plain old buffered or un buffered memory to avoid the problems of the random Nmi error messages in windows, with a stop blue screen error.

 

[Tradesman1]

Would help to know the hardware, DRAM model #(s), CPU mobo, do you have latest BIOS, is memory running at spec, is the CPU (it's MC) capable of carrying the DRAM. there's all kinds of things that can adversely affect DRAM that is perfectly good

 

[lightingft]

If you are getting Nmi parity errors.

You need to check your memory sticks visually.
look on the sticker or label of each memory module you have.

If it states the memory is Ecc, then you need to enable the Ecc option in your bios of the motherboard if it exists.

If the memory does not state it is Ecc enabled on the sticker you need to again go into the bios and disable the Ecc option in the bios of the motherboard.

Nmi is an extra parity bit error, so it suggests that the memory you have is Ecc capable.
And the option in your bios is not currently enabled in the bios.

Ecc enabled memory does not like it if the motherboard bios does not support the feature for the memory to enable the extra parity bit it uses. And will throw up the Nmi error from time to time within windows.

If you find it to be the case then you will have to change your memory to plain old buffered or un buffered memory to avoid the problems of the random Nmi error messages in windows, with a stop blue screen error.

Thank you, I will try to check into it. My RAM stick doesn't have anything like that written on it.
EDIT: Nope, that is not the case. I forgot to mention I am on a desktop, so neither my RAM is ECC, and my mobo doesn't even support it I think.

Would help to know the hardware, DRAM model #(s), CPU mobo, do you have latest BIOS, is memory running at spec, is the CPU (it's MC) capable of carrying the DRAM. there's all kinds of things that can adversely affect DRAM that is perfectly good

PSU: Corsair RM750i
Motherboard: ASRock Z75 Pro3
CPU: Intel i7-2600 3.4GHz
Video Card: MSI Nvidia GTX 970 Gaming 4G
RAM: 8GB DDR3
I haven't done anything to BIOS after installing the MOBO, but yes, it's the latest. 2.0.
Pretty sure the CPU is capable, yes.
Both RAM sticks are Kingston, KVR1333D3N9/4G.

When you say "passed all MemTests" do you mean that you ran MemTest86 for several clean passes?

Ah yes, I forgot to mention. I've left it for about a night and I think there were about 8~ clean passes.


Also, one more thing. The NMI parity errors are very rare. Usually just a lock up.

 

[Tradesman1]

If DRAM is running at spec, I'd guess more likely a driver or hard drive issue. Have you run system file checker or the diskcheck?

Go to command prompt as Admin and run the file checker - command is SFC /SCANNOW

the disk check takes forever so I run overnight - same as a bove from command prompt the command is CHKDSK /F /R

 

[lightingft]

If DRAM is running at spec, I'd guess more likely a driver or hard drive issue. Have you run system file checker or the diskcheck?

Go to command prompt as Admin and run the file checker - command is SFC /SCANNOW

the disk check takes forever so I run overnight - same as a bove from command prompt the command is CHKDSK /F /R

Alright, thank you, I'll try this. There shouldn't be hard drive issues though. I currently only have an SSD and a 1 year old hard drive. I removed a 5 year old hard drive that was pretty much dying.

 

[Kewlx25]

I've seen memory that took nearly a full week of memtest to find the error.

 

[lightingft]

If DRAM is running at spec, I'd guess more likely a driver or hard drive issue. Have you run system file checker or the diskcheck?

Go to command prompt as Admin and run the file checker - command is SFC /SCANNOW

the disk check takes forever so I run overnight - same as a bove from command prompt the command is CHKDSK /F /R

My PC just froze and screen went black. Could it be a GPU issue? My GPU is only a few months old though, is this behaviour possible if it's loose in the mobo? Although I doubt that's the issue, I've reseated it alot of times.

 

[Tradesman1]

It could be a lot of things, could try reseating the GPU

 

[lightingft]

It could be a lot of things, could try reseating the GPU

I've pretty much slimmed out the problems though.. Alot of my PC hardware was replaced very recently. The only things that haven't are RAM, one HDD and the CPU. I can only suspect the RAM, drivers or the GPU right now.. I'll try doing a /SFC scannow and doing a longer MemTest for the time being.. But the crashes haven't happened at one point for 12 days which is weird. This problem is not consistent at all, but seems to affect me mostly at evening-night (around 9PM-1AM).

 

[lightingft]

If DRAM is running at spec, I'd guess more likely a driver or hard drive issue. Have you run system file checker or the diskcheck?

Go to command prompt as Admin and run the file checker - command is SFC /SCANNOW

the disk check takes forever so I run overnight - same as a bove from command prompt the command is CHKDSK /F /R

Oh yes, now I remember... When I tried to SFC /SCANNOW there were alot of errors that couldn't be fixed. I tried to fix it with Windows startup disk and CMD prompt there but I had the same issue. What should I do? I've never got to fixing those errors.
It says:
'Windows Resource Protector found corrupt files but was unable to fix some of them' and a log was created.
Also, I'm on Windows 7.

 

[Tradesman1]

Not sure if available, but I normally list the bad files and get good copies off another rig and replace them - if you have access to another rig with the same OS

 

[lightingft]

Not sure if available, but I normally list the bad files and get good copies off another rig and replace them - if you have access to another rig with the same OS

I'm not sure I have such access, any other way to fix these files?