help in interpreting the results

[mikerzepa]

My kid was complaining about computer crashes during a game. I tested that computer (a new computer) using memtest86. I got thousands of errors and a test is also crashing so it can't be completed. Can someone take a look at the pictures and tell me what is most probably broken, is it a CPU or it can be a motherboard?

[Richard (PassMark)]

Look like bad RAM.

If you have multiple RAM sticks, you can try testing them individually, else try swapping the RAM with another set. More information on troubleshooting can be found here,

MemTest86 - Troubleshooting Memory Errors

[mikerzepa]

Look like bad RAM.

If you have multiple RAM sticks, you can try testing them individually, else try swapping the RAM with another set. More information on troubleshooting can be found here,

MemTest86 - Troubleshooting Memory Errors

I forgot to mention that I've replaced ram already.

[David (PassMark)]

So were the screen shots from the original RAM or the new replacement RAM?

[mikerzepa]

So were the screen shot from the original RAM or the new replacement RAM?

screens are from new RAM. Old RAM was also giving thousands of errors in test.

[David (PassMark)]

If it is a new computer under warranty, send it back for repair.

If you built it from parts, you can try testing them individually in different slots, as it might be a bad slot.

[mikerzepa]

If it is a new computer under warranty, send it back for repair.

If you built it from parts, you can try testing them individually in different slots, as it might be a bad slot.

I did what you said - I made 4 tests with a single memory module: module1 in bank1, module1 in bank2, module2 in bank1, module2 in bank2. Every attempt was finished with test failure because of too many errors.

What is most probable reason - CPU or motherboard?

[David (PassMark)]

More chance it is the CPU that the motherboard. But could be either.

Also was the replacement RAM identical (model and part number) to the original?
As it could be a BIOS bug. Setting the wrong clock speeds, timings or voltage levels. Check if there is new BIOS firmware for the motherboard.

[mikerzepa]

More chance it is the CPU that the motherboard. But could be either.

Also was the replacement RAM identical (model and part number) to the original?
As it could be a BIOS bug. Setting the wrong clock speeds, timings or voltage levels. Check if there is new BIOS firmware for the motherboard.

I made my MB's (Gigabyte B560M H) BIOS upgrade to the newest version (F9) yesterday. I noticed that all BIOS settings were restored to default.
I ran memTest and got no errors but after I set XMP 3200 (which was set before BIOS upgrade) I got tons of errors again.

Finally I set XMP 3000 and ran memTest without any errors.
What I can do to make my set up working stable on XMP 3200 setting?
Should I claim my MB under warranty?

Memory modules are 2 x Kingston FURY Renegade 8GB
CPU is Intel Core i5-11400F

[mikerzepa]

memory P/N is KF3200C16D4/8GX and I didn't find it in Gigabyte B560M H supported memory list

[David (PassMark)]

and I didn't find it in Gigabyte B560M H supported memory list

And I would expect that to be the answer you get from Gigabyte as well.

So you have likely be sold products that don't meet their specs. But hard to know for sure who is to blame without a lot of testing, or using some every high end test equipment.
On the other hand you won't notice the performance different between 3000 and 3200. It will be so small that you'll have trouble measuring the difference. Especially if the latency timings are tighter at the slower clock speed.

You might find you can manually edit the settings in BIOS to get it stable at the faster clock speed (voltages and timings), But honestly you shouldn't have to and it isn't worth the effort generally.