Possible memory problems. How can I be certain?

X82

X82

Associate
Joined
24 May 2010
Posts
212
For years now, my computer has always crashed while playing games. I used to brush it off but over the years I've slowly become tired and dare I say it, even depressed over my computers ability to crash.

First the specs:

Operating System
Windows 10 Pro 64-bit

CPU
Intel Core i7 2600K @ 3.40GHz 35 °C
Sandy Bridge 32nm Technology

RAM
16.0GB Dual-Channel DDR3 @ 668MHz (9-9-9-24)

Motherboard
ASUSTeK Computer INC. P8Z68-V LX (LGA1155) 29 °C

Graphics
3071MB NVIDIA GeForce GTX 1060 3GB (Gigabyte) 57 °C

Storage
931GB Western Digital WDC WD10EACS-00ZJB0 (SATA ) 32 °C
0B SAMSUNG HD204UI (SATA )
931GB SAMSUNG HD103SI (SATA ) 24 °C
238GB M4-CT256M4SSD2 (SATA (SSD))

PSU
Unknown model, but it is a Corsair 1000W


I've had the occasional bluescreen when not gaming, some cases when the machine is idle, but the pattern seems clear.
Games while crash to desktop with little warning and only in rare cases display any error message.

When they do however, it is nearly always an "Access Violation" message. In come cases after the game crashes the system then bluescreens also.

Here is a folder of all the dump files I have on file plus some other files.

Favourite bluescreen hits include:

IRQL_NOT_LESS_OR_EQUAL

MEMORY_MANAGEMENT

SYSTEM_SERVICE_EXCEPTION


plus many more.

With all the errors and other posts I've made, a general consensus is that it is my memory which is the issue. However Memtest86 ran with all default, 4 passes all night with no issues.

That being said, I ran it again and when I woke up I noticed it had frozen 2 minutes after I left the room.
I then ran it once more last night and this time it went for nearly 4 hours, all tests passed.

However, I ran Memtesk HCI and that threw errors up everywhere.

I'm saving up to be able to buy replacement memory at the end of the month. Found a good deal I think, but I don't want to waste money on RAM when that isn't the issue.

How can I be truly certain the memory is borked? Is memtest86 the only true way?
If it is the PSU, how can I even test that to confirm it?
 
Last edited:
Soldato
Joined
27 Mar 2009
Posts
3,301
Just a heads up you should remove that hyperlink to the other shop selling RAM, it's against forum rules.

With your problem it's unlikely that both RAM sticks you have are faulty so it's best to test one at a time with your CPU at stock.

This way if you get errors with both sticks individually you can start looking at it being something else.

Then if it is just one stick depending on the make there is a chance it could still be in warranty and you could get a replacement.

Just test with HCI memtest if that is throwing errors the quickest. Saves you hours of running other tests
 

X82

X82

Associate
OP
Joined
24 May 2010
Posts
212
Thanks, removed link.
So its possible that memtesk86 which people rave about, could actually just not show errors?

I will re-run HCI memtest again. But when I ran it before, I asked me to run multiple instances and allocate memory to each one. I think I did 3GB per instance. I don't think there was a way to select specific sticks, unless I missed a setting.
 
Associate
Joined
23 Dec 2018
Posts
1,102
I had a similar issue years back with Memtest not showing errors even after about 5 overnight tests, but it finally showed some on a final run before I got new ram that was 100% stable for years in the same system, with no BSODs subsequently, so it can take a while for them to show.

Those BSOD codes seem like ram issues as well, pretty sure I had similar ones and also the random nature of the crashes, often gaming but also randomly in desktop, etc.

Hope you get it sorted out.
 
Soldato
Joined
27 Mar 2009
Posts
3,301
With HCI memtest you should run one instance for every CPU thread you have so in your case 8.

You are going to have to physically remove one stick of RAM, test, then swap over etc...
 

X82

X82

Associate
OP
Joined
24 May 2010
Posts
212
I had a similar issue years back with Memtest not showing errors even after about 5 overnight tests, but it finally showed some on a final run before I got new ram that was 100% stable for years in the same system, with no BSODs subsequently, so it can take a while for them to show.

Did you have issues in games previously? I'm hoping that if I get new RAM these problems will stop completely.

With HCI memtest you should run one instance for every CPU thread you have so in your case 8.
You are going to have to physically remove one stick of RAM, test, then swap over etc...

Is there any issues running this with all sticks in?
I ask because even if one stick was faulty, the cost to replace it is near the price for entirely new RAM.
 
Soldato
Joined
27 Mar 2009
Posts
3,301
Is there any issues running this with all sticks in?
I ask because even if one stick was faulty, the cost to replace it is near the price for entirely new RAM.

No issues it just doesn't help you narrow down the issue. If you test both sticks individually and they both fail you can start to look at CPU settings or RAM slots as it's unlikely both would be bad.

I've also had it a couple of times on 1155 systems like yours having memory crashes because the CPU cooler was overtightened and distorting the board itself
 

X82

X82

Associate
OP
Joined
24 May 2010
Posts
212
Welp.

I ran the HCI test on each stick in turn in safe mode.
All passed. I then tried one stick in each slot and ran each test again.
Passed.
I then rebooted with all sticks in all slots and ran again, passed.

I am actually annoyed. I was hoping for a bad stick or two. Hell, even a bad slot. I could survive on less ram. But this is back to square one and I can't fork out £70 on memory if I don't need it.

Also, might be unrelated. But in Windows, it shows my RAM running at a speed of 1333mhz
But the ram is 1600mhz. Is this normal?
 
Last edited:
Soldato
Joined
27 Mar 2009
Posts
3,301
Welp.

I ran the HCI test on each stick in turn in safe mode.
All passed. I then tried one stick in each slot and ran each test again.
Passed.
I then rebooted with all sticks in all slots and ran again, passed.

I am actually annoyed. I was hoping for a bad stick or two. Hell, even a bad slot. I could survive on less ram. But this is back to square one and I can't fork out £70 on memory if I don't need it.

Also, might be unrelated. But in Windows, it shows my RAM running at a speed of 1333mhz
But the ram is 1600mhz. Is this normal?

nope, Have you enabled the XMP profile in BIOS?

Just checked your mobo manual, make sure the RAM are in slots 2+4 (counting left to right) these are the primary slots
 

X82

X82

Associate
OP
Joined
24 May 2010
Posts
212
So ~8 years I've had this motherboard and memory. In all that time I never even knew of that setting.
Just activated it and windows shows the correct speed. Now, I don't know what gains/losses will come as a result of it, but still. The more you know.

I will try some more games to see if I can cause another crash, as I wonder if reseating the ram helped at all.
 
Back
Top Bottom