Same errors with 2 different ram modules

[aquarius]

Hi everyone!
I made a Memtest86 7.3 UEFI test on my notebook Asus X553MA (updated to latest bios) with the original 4GB DDR3 SK Hynix HMT451S6BFR8A-PB, and it found errors on:

Test 3 > Errors 1
Test 4 > Errors 1
Test 5 > Errors 1
Test 7 > Errors 1
Test 8 > Errors 1
Test 9 > Errors 1
Test 11 > Errors 1
Test 12 > Errors 1
Test 13 > Errors 6

Test 0, 1, 2, 6 and 10 are OK.

So, I change the Hynix module with a brand new Kingston 4GB DDR3 KVR16LS11/4 and it found the SAME errors on the SAME tests.

My questions are:
1. is there any incompatibility with that Asus motherboard ??
2. i know that Memtest86 7.3 is better and updated than Memtest86+ 5.01 version, but why the last one found NO ERRORS at all??


Thanks in advance.

[David (PassMark)]

Can you post the details of the errors. In particular is the error address the same for all tests (and for the 2 sticks of RAM).

There are known issues with ASUS and the multi-threading option (due to BIOS bugs). But that generally results in a crash or a freeze, not a RAM error.

MemTest86+ V5 doesn't use UEFI. So the range of memory addresses tested in MemTest86 V7 in UEFI in different from the older non UEFI releases.

[aquarius]

I post you the errors from Test 13 which got six errors, but as you

These are from Kingston 4GB DDR3

[Data Error] Test: 13, CPU: 0, Address: 0, Expected: 3A50C3A2, Actual: 3A50C301

[Data Error] Test: 13, CPU: 0, Address: 0, Expected: C03FAA15, Actual: C03FAA01

[Data Error] Test: 13, CPU: 0, Address: 0, Expected: 85929AFB, Actual: 85929A01

[Data Error] Test: 13, CPU: 0, Address: 0, Expected: CC8C1ABD, Actual: CC8C1A01

[Data Error] Test: 13, CPU: 0, Address: 0, Expected: 49A7ABE4, Actual: 49A7AB01

[Data Error] Test: 13, CPU: 0, Address: 0, Expected: EBE5980F, Actual: EBE59801

These are from Hynix 4GB DDR3

[Data Error] Test: 13, CPU: 0, Address: 0, Expected: 612237A2, Actual: 61223701

[Data Error] Test: 13, CPU: 0, Address: 0, Expected: 80701215, Actual: 80701201

[Data Error] Test: 13, CPU: 0, Address: 0, Expected: C24066FB, Actual: C2406601

[Data Error] Test: 13, CPU: 0, Address: 0, Expected: 40C4CEBD, Actual: 40C4CE01

[Data Error] Test: 13, CPU: 0, Address: 0, Expected: 386B43E4, Actual: 386B4301

[Data Error] Test: 13, CPU: 0, Address: 0, Expected: 8A60B40F, Actual: 8A60B401

Well, as you may notice the errors are not exactly the same as I said before, sorry, I only watched "Test number" errors.
Second, I think you got the point, because doing again tests on Kingston and Hynix setting ONLY ONE Cpu, no more errors.

Hope this help you.

[David (PassMark)]

From the information you provided it is inconclusive.
No errors when testing with 1 CPU might mean that the RAM doesn't get errors until it is hot. Or that there is a bad CPU core,

[keith]

The errors are from Test 13, which indicate that your RAM is susceptible to row hammer errors. See this page for details:

http://www.memtest86.com/troubleshooting.htm#hammer

[aquarius]

From the information you provided it is inconclusive.
No errors when testing with 1 CPU might mean that the RAM doesn't get errors until it is hot. Or that there is a bad CPU core,

wait... how can it be "hot" ?? the first error found is on Test #3, I think nothing could be "hot" so soon.
could it be an issue between Asus motherboard and multi-threading as you said before??

@ keith
if you look up at my first message you will see that not only Test #13 got errors, but many other Tests.

[keith]

Try turning on CSM (Compatibility Support Module) in the BIOS setup.

See this thread:
http://www.passmark.com/forum/memtes...pu-cores/page3