Strange(?) Memory Errors

[taragor]

Hello Everyone,

so i've just switched to a new Mainboard and had some System instabilitys since then. To Check memory I ran MemTest86 and indeed got some (6) errors in 4 passes.
Since all were on the same address i guessed i had a dead Stick of RAM, so i tested both sticks seperately. Both sticks passed 5 tests each without any error.
So I guessed it was a CPU/MB error and tested all Slots seperately (only one occupied at a time). Again all slots passed 5 passes without any errors. After that i tested both Sticks on the same Channel (Single Channel mode) and again got 0 errors on 5 passes. Since windows was working also in this setting I reset the BIOS again (even reflashed it) and ran a new test in DualChannel mode after making sure the Memory specs are correct (2400MHz, correct timings, 1.2V). The test is still running but until now i have 3 errors within the first 2 passes, with the first one finishing without any errors, one error occuring in test 5-pass 2 and the others in hammers test pass 2. I will upload the log once the test finishes.
By now I'm very confused by the results, since I both sticks seem to be alright (I guess?) and all my channels seem to be fine to. Yet i keep getting errors in dual channel mode.
What do you guys make out of all that?
Am I looking on a (slightly?) bad stick of RAM or is it more a CPU/MOBO sided error?

Also i found that my Memory isn't explicitly supported by the Motherboard, accrding to the support list. Might that be a cause for those errors?

My Specs are:
Intel i5-8400
Gigabyte Aorus Z370 Gaming 7
2x8Gb RAM 2400MHz, 16-16-16-39, 1.2V (Team Group)
Asus Strix RTX 2070 Advanced

[David (PassMark)]

Am I looking on a (slightly?) bad stick of RAM

Yes this is what I would suspect. It can happen that RAM can pass in single channel mode but fail in dual channel mode. The access pattern is different.

PC Memory is standardised. So in theory at least it doesn't matter which RAM you use with which motherboard as long as the basic specs are meet (DDR3, DDR4, etc..). In the real world there are cases of some motherboards and RAM being marginal when used in combination.

[taragor]

Thanks for your answer.
Since the errors were all located in a very narrow field of addresses (within 32MB) and all had the same Error mask I came to the same conclusion.
What I still think is odd is the fact, that all the errors apparently come from the same core (core 4) in all my tests. Even stranger is that using only one core the test passes if I use any core but core 4.
However it seems unlikely that a single core, that doesn't even communicate with the ram directly but through a memory controller, develops a habit of getting one (always the same) faulty bit within a that small range of addresses.
Or am I getting this wrong?

I've added a log of a test for reference.

For now i've ordered new RAM and hope this will solve this problem.
Summary

Report Date	2019-04-01 22:13:45
Generated by	MemTest86 V8.1 Free (64-bit)
Result	FAIL

System Information

EFI Specifications	2.60
System
Manufacturer	Gigabyte Technology Co., Ltd.
Product Name	Z370 AORUS Gaming 7-OP
Version	Default string
Serial Number	Default string
BIOS
Vendor	American Megatrends Inc.
Version	F13
Release Date	03/22/2019
Baseboard
Manufacturer	Gigabyte Technology Co., Ltd.
Product Name	Z370 AORUS Gaming 7-OP
Version	Default string
Serial Number	Default string
CPU Type	Intel Core i5-8400 @ 2.80GHz
CPU Clock	2837 MHz [Turbo: 3910.6 MHz]
# Logical Processors	6
L1 Cache	4 x 64K (243775 MB/s)
L2 Cache	4 x 256K (92463 MB/s)
L3 Cache	9216K (26755 MB/s)
Memory	16339M (17755 MB/s)
DIMM Slot #0	8GB DDR4 PC4-19200
	Team Group Inc. / TEAMGROUP-UD4-2400 / 01047CE8
	16-16-16-39 / 2400 MHz / 1.2V
DIMM Slot #1	8GB DDR4 PC4-19200
	Team Group Inc. / TEAMGROUP-UD4-2400 / 0104C112
	16-16-16-39 / 2400 MHz / 1.2V

Result summary

Test Start Time	2019-04-01 22:06:34
Elapsed Time	0:07:07
Memory Range Tested	0x0 - 43F000000 (17392MB)
CPU Selection Mode	Parallel (All CPUs)
ECC Polling	Enabled
# Tests Passed	5/7 (71%)

Lowest Error Address	0x3F4E091C8 (16206MB)
Highest Error Address	0x3F6E093C8 (16238MB)
Bits in Error Mask	0000000000000100
Bits in Error	1
Max Contiguous Errors	1

Test	# Tests Passed	Errors
Test 0 [Address test, walking ones, 1 CPU]	1/1 (100%)	0
Test 1 [Address test, own address, 1 CPU]	1/1 (100%)	0
Test 2 [Address test, own address]	1/1 (100%)	0
Test 3 [Moving inversions, ones & zeroes]	1/1 (100%)	0
Test 4 [Moving inversions, 8-bit pattern]	1/1 (100%)	0
Test 5 [Moving inversions, random pattern]	0/1 (0%)	1
Test 6 [Block move, 64-byte blocks]	0/1 (0%)	4
Test 7 [Moving inversions, 32-bit pattern]	0/0 (0%)	0
Test 8 [Random number sequence]	0/0 (0%)	0
Test 9 [Modulo 20, ones & zeros]	0/0 (0%)	0
Test 10 [Bit fade test, 2 patterns, 1 CPU]	0/0 (0%)	0
Test 13 [Hammer test]	0/0 (0%)	0

Last 10 Errors

2019-04-01 22:10:31 - [Data Error] Test: 6, CPU: 4, Address: 3F6E093C8, Expected: 00000400, Actual: 00000500

2019-04-01 22:10:31 - [Data Error] Test: 6, CPU: 4, Address: 3F6E09388, Expected: 00000200, Actual: 00000300

2019-04-01 22:10:31 - [Data Error] Test: 6, CPU: 4, Address: 3F4E093E8, Expected: 00000400, Actual: 00000500

2019-04-01 22:10:31 - [Data Error] Test: 6, CPU: 4, Address: 3F4E093A8, Expected: 00000200, Actual: 00000300

2019-04-01 22:09:38 - [Data Error] Test: 5, CPU: 4, Address: 3F4E091C8, Expected: 9C29A885, Actual: 9C29A985

Certification

This document certifies that the Tests described above have been carried out by a suitably qualified technician on the System described above.

Signed

___________________________

[David (PassMark)]

Real test is to replace the RAM and see if the problem goes away.

[taragor]

Thanks for your great help.
Swapped to new RAM and all is working again, no Errors anywhere.