F3-12800CL10S-8GBXL Memtest errors

[qiqi1021]

Hi,

I bought 4 of these 8GB sticks to upgrade my 2 computers (2x8GB in each PC). Out of the 4 sticks I purchased 3 are working perfectly but one seems to be faulty.

I have Memtested that stick run at 1600MHz 10-10-10-30-2T timings as per G.Skill specifications programmed in XMP, as well as standard SPD specs 1333MHz 9-9-9-24-2T. Both settings come up with errors. Errors pop up in both memory slots A1 and B1 at 1.5V in a Gigabyte GA-P67A-UD4-B3 with i5 2320. All three other sticks tested in the same configuration without a problem.

Pictures:



Strangely, when I move this bad stick to my other PC (i5 2500K, Asus P8Z68-V Pro) Memtest passes but is unstable running IntelBurnTest in Windows. The three other sticks fly through IntelBurnTest in this PC without a hitch.

Do I have a bad module? Thanks.

[Britton30]

Something to try is to set your DRAM frequency, Timings, and voltage manually. This may work.

[qiqi1021]

The frequency, main timings, command and voltage were all set manually, I only let the subtimings run on auto.

[Tradesman]

May be a bad stick, I'd look to return/exhange or RMA

[qiqi1021]

Thanks for the confirmation. The retailer took a look at the photos and came to the same conclusion. Will be sending it back for an exchange.

[Tradesman]

Let us know when you get the new stick

[qiqi1021]

I've received the replacement module but now found one of the other originals seems to be problematic as well.

First, some more background info. I originally received 4 modules with serial numbers
12261600271949
12261600271950
12261600271954 which was the faulty one
12261600271955

First thing I did was run Memtest86+ v4.20 on each stick individually in my Gigabyte P67A-UD4-B3.

After passing (and failing) the sticks ending in 49 and 50 went into my Asus P8Z68-V Pro, where they passed all FFT lengths in Prime95 v27.7 custom blend 14336MB (14GB) without error (that's about 21 hours).

The sticks ending in 54 and 55 were left for the P67-UD4-B3, awaiting a replacement for 54. I didn't do any further tests on 55 since I thought passing Memtest should be fine...

I received the replacement module numbered 12261600271953 on Thursday, set it up with 55 in the P67A-UD4-B3 and the 2x8GB ran through 6 whole passes of Memtest86+ v4.20 in 14 hours without error.

Then I started up P95 and came back 4 hours later to see one thread errored. Messed with various combinations of VCCIO and VCCSA and they still error in P95 even with both set to 1.10V. I thought maybe these modules just don't like the UD4 and moved them into my P8Z68-V Pro.

And... same thing, errors in P95. Decided to test each stick individually, 53 successfully passed all P95 FFT lengths testing 6144MB (6GB) custom blend without error.

Put 55 in and found it errors consistently around the 6 minute mark. CPU's at stock clock, all voltages at stock, and the same settings passed P95 with the 49, 50 and 53 sticks so it definitely is a problem with this module. I've tried raising VDIMM incrementally up to 1.55V but it still errors, hope that doesn't void warranty? Last settings tested was 1.10V VCCIO 1.55V VDIMM and errored too (no VCCSA setting on the P8Z68-V Pro). Is this another bad stick?

Both my motherboards are definitely fine, thoroughly stress tested from day 1. The P67 board has been used to Memtest all 5 sticks so it's definitely not randomly killing RAM. It's currently running a 2x4GB kit from a competitor (and F3-12800CL9D-8GBRL pior to that) without problem. Will try the 49 and 50 modules in the UD4 tomorrow.

[Tradesman]

So you RMAed a single stick out of a set (set was 54-55?) and now have a 53 and 55, correct? or was this 4 individual sticks, each in own package? This sounds odd, let me see if GSkill or GSkill Tech can take a look, think GSkill will be in this evening, GSkill Tech is more M-F my time. If they were individual sticks and were tested together it may have produced errors, did you test individually? If individual sticks, may have ones that don't want to play together, which is why GSkill always suggest buying a single set, XMP won't really work with mixed packaged so need to set up manually and often requires additional voltage.

[qiqi1021]

All modules are individual sticks each in own package.

In the P8Z68-V Pro the 49 and 50 sticks passed P95 in dual channel.

53 and 55 didn't, then tested individually and found 53 passes on its own. 55 consistently fails around 6 minutes in regardless of VDIMM and VCCIO settings. All frequencies and voltages stock, XMP enabled with main timings manually set 10-10-10-30-2T, subtimings on auto. The only thing changed between tests was the RAM module.

Now that I think about it, I will try XMP disabled with frequency and timings manually set tonight and see how that goes. Thanks.

[Tradesman]

Give it a try and let us know, with mixed sets (or individual sticks) may need a litle extra DRAM and/or VCCIO voltage

[qiqi1021]

Testing just the 53 stick, I've tried every combination of 1.50/1.55/1.58V VDIMM and 1.05/1.075/1.10V VCCIO and still failing P95 with XMP off. XMP on appeared less stable but the P8Z68-V Pro reports the exact same subtimings in the UEFI and in Aida64 in Windows as XMP off.

Most stable I got was with XMP off, VDIMM 1.50V and VCCIO 1.10V, and even then it only ran for 38 minutes before erroring.

Max voltage I tested was 1.60V VDIMM and 1.12V VCCIO, same result.

There's no VCCSA setting on the P8Z68-V Pro but I remember reading somewhere it's automatically set by the motherboard to the same voltage as the VCCIO setting.

Here's my P95 log...

[Sun Jul 22 23:22:33 2012]
FATAL ERROR: Rounding was 0.49609375, expected less than 0.4
Hardware failure detected, consult stress.txt file.
[Sun Jul 22 23:30:36 2012]
Self-test 448K passed!
Self-test 448K passed!
Self-test 448K passed!
[Sun Jul 22 23:52:08 2012]
FATAL ERROR: Rounding was 0.498046875, expected less than 0.4
Hardware failure detected, consult stress.txt file.
[Mon Jul 23 00:00:27 2012]
Self-test 448K passed!
Self-test 448K passed!
Self-test 448K passed!
[Mon Jul 23 00:23:35 2012]
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
[Mon Jul 23 17:15:07 2012]
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
[Mon Jul 23 17:20:14 2012]
FATAL ERROR: Rounding was 0.4995117188, expected less than 0.4
Hardware failure detected, consult stress.txt file.
[Mon Jul 23 17:32:46 2012]
Self-test 448K passed!
Self-test 448K passed!
Self-test 448K passed!
[Mon Jul 23 17:53:45 2012]
Self-test 448K passed!
Self-test 448K passed!
Self-test 448K passed!
Self-test 448K passed!
[Mon Jul 23 18:17:39 2012]
FATAL ERROR: Rounding was 0.4992675781, expected less than 0.4
Hardware failure detected, consult stress.txt file.
Self-test 448K passed!
Self-test 448K passed!
Self-test 448K passed!
[Mon Jul 23 18:46:40 2012]
Self-test 448K passed!
Self-test 448K passed!
Self-test 448K passed!
Self-test 448K passed!
[Mon Jul 23 19:01:47 2012]
Self-test 8K passed!
Self-test 8K passed!
Self-test 8K passed!
Self-test 8K passed!
[Mon Jul 23 19:21:26 2012]
FATAL ERROR: Rounding was 0.4997558594, expected less than 0.4
Hardware failure detected, consult stress.txt file.
Self-test 448K passed!
Self-test 448K passed!
Self-test 448K passed!
[Mon Jul 23 19:51:36 2012]
Self-test 448K passed!
Self-test 448K passed!
Self-test 448K passed!
Self-test 448K passed!
[Mon Jul 23 20:06:40 2012]
Self-test 8K passed!
Self-test 8K passed!
Self-test 8K passed!
Self-test 8K passed!
[Mon Jul 23 20:14:57 2012]
FATAL ERROR: Rounding was 0.4892578125, expected less than 0.4
Hardware failure detected, consult stress.txt file.
[Mon Jul 23 20:22:20 2012]
Self-test 512K passed!
Self-test 512K passed!
Self-test 512K passed!
[Mon Jul 23 20:37:42 2012]
Self-test 12K passed!
Self-test 12K passed!
Self-test 12K passed!
[Mon Jul 23 20:52:46 2012]
Self-test 576K passed!
Self-test 576K passed!
Self-test 576K passed!
[Mon Jul 23 21:08:06 2012]
Self-test 18K passed!
Self-test 18K passed!
Self-test 18K passed!
[Mon Jul 23 21:23:22 2012]
Self-test 672K passed!
Self-test 672K passed!
Self-test 672K passed!
[Mon Jul 23 21:39:23 2012]
Self-test 21K passed!
Self-test 21K passed!
Self-test 21K passed!
[Mon Jul 23 21:54:35 2012]
Self-test 768K passed!
Self-test 768K passed!
Self-test 768K passed!
[Mon Jul 23 23:10:42 2012]
Self-test 448K passed!
Self-test 448K passed!
Self-test 448K passed!
Self-test 448K passed!
[Mon Jul 23 23:26:11 2012]
Self-test 8K passed!
Self-test 8K passed!
Self-test 8K passed!
Self-test 8K passed!
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
[Tue Jul 24 13:18:58 2012]
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
[Tue Jul 24 13:39:07 2012]
Self-test 448K passed!
Self-test 448K passed!
Self-test 448K passed!
Self-test 448K passed!
[Tue Jul 24 13:54:18 2012]
Self-test 8K passed!
Self-test 8K passed!
Self-test 8K passed!
Self-test 8K passed!
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
[Tue Jul 24 14:09:29 2012]
Self-test 512K passed!
Self-test 512K passed!
Self-test 512K passed!

[Tradesman]

Sounds like another bad stick, think I'd RMA

[qiqi1021]

Yes will do, I'm very disappointed. This is only the third time I've purchased G.Skill and the first time I've received RAM DOA.

[Tradesman]

Think I've only had 1 out of probably a couple hundred or more sets

[qiqi1021]

Got the second replacement and it works fine. It runs in dual channel with one of the other modules all on stock voltages even though it's manufactured a month prior, they've blasted through 22 hours of P95 14GB blend without breaking a sweat.

So basically, somehow 2 consecutive bad sticks managed to sneak through G.Skill QC...