How to relate to errors in Hammer Test 13 ?

[Bleep]

@Bleep

Cannot give you any guarantee. But your symptoms looks more like unstable GPU/drivers combination. Been a lot of Chrome/NVIDIA references lately (if that is your browser/GPU). Or a slightly unstable system because of some overclock/tune/BIOS of CPU/RAM/GPU.

I wouldn't rule out the RAM sticks. But I haven't read much about normal PC's being that unstable with only a few hammer 13 errors. Although hammer 13 are real errors, the memory usage to provoke them is not very likely with normal PC usage.

My advice is that you look at your freeze problem from more angles than just memory. Independent of that, your hammer 13 errors are real. As you have read in this thread, no easy answer. David's Troubleshooting Memory Errors link contains good solid advices for memory and hammer trouble in general.

When it comes to faulty/refund. If you still get hammer 13 errors when testing the RAM sticks individually, at stock spec. Yes, then I would say you have a good argument to get them replaced. These are real errors that shouldn't be there.

I reckon you're right there. I'm using my motherboard's built-in video (Intel HD 4600) so that may be why the crash was so fatal and showed no errors in Event Viewer. I've updated my BIOS which mentioned "Adjust shared memory setting" and "Update VBIOS" so that sounds promising. Now I have to try and make it crash again.

I've read the Troubleshooting Memory Errors and this entire thread (which I'll keep reading as my inner nerd compels me) and I'll keep an eye out for any glitches. I'm sure the freezing is a hardware issue and finding the cause won't be easy.

Nothing is overclocked nor has it or will it ever be. I like a predictable, reliable system

[sveinan]

@Bleep

Cannot give you any guarantee. But your symptoms looks more like unstable GPU/drivers combination. Been a lot of Chrome/NVIDIA references lately (if that is your browser/GPU). Or a slightly unstable system because of some overclock/tune/BIOS of CPU/RAM/GPU.

I wouldn't rule out the RAM sticks. But I haven't read much about normal PC's being that unstable with only a few hammer 13 errors. Although hammer 13 are real errors, the memory usage to provoke them is not very likely with normal PC usage.

My advice is that you look at your freeze problem from more angles than just memory. Independent of that, your hammer 13 errors are real. As you have read in this thread, no easy answer. David's Troubleshooting Memory Errors link contains good solid advices for memory and hammer trouble in general.

When it comes to faulty/refund. If you still get hammer 13 errors when testing the RAM sticks individually, at stock spec. Yes, then I would say you have a good argument to get them replaced. These are real errors that shouldn't be there.

[Bleep]

Bleep, see this page to start with. Troubleshooting Memory Errors.

It might not have frozen. The test is a bit slow at the moment and the progress bar doesn't update smoothly. We are hoping to improve this soon in a V6.1 release. (Probably next week).

My bad, to clarify I was playing a game when the system froze (looping sound, mouse and keyboard didn't work, no event logged at the time the system froze and the PC restarted after 20 or 30 seconds). The system also half froze a few days later, but it came back after five or so seconds and hasn't had an issue since.

[David (PassMark)]

Bleep, see this page to start with. Troubleshooting Memory Errors.

It might not have frozen. The test is a bit slow at the moment and the progress bar doesn't update smoothly. We are hoping to improve this soon in a V6.1 release. (Probably next week).

[Bleep]

As part of trying to find out why my newly built system froze while I was playing a game and partly froze while I was web browsing, I ran memtest86 6.0 and I have three errors right near the end of the hammer test. Am I right in thinking the RAM is faulty and I can return it for something better? I found the Kingston HyperX Savage locally (Australia) and if I can get a refund on the G-Skill I'll grab some.

The RAM I tested with is:
G-Skill Ripjaws X 8GB (2x4GB)
12800CL9D-8GBXL
DDR3 1600 CL9-9-9-24 1.50v
PC3-12800 4Gx2 intel XMP Ready

[David (PassMark)]

The might have increased the RAM refresh rate via a change BIOS to reduce the probability of row hammer errors. This increases power usage however, which isn't really desirable for a laptop.

[sveinan]

A quick FYI (BIOS update with security mitigation for row hammer),

As mentioned earlier my laptop is a ThinkPad X230 (which gets 1 hammer error, sometimes). Was going through a refresh of drivers on it, also found an updated BIOS (05may2015). Support page on Lenovo can be found here. Newest BIOS 2.64 README, can be found here. Quote from README:

CHANGES IN THIS RELEASE
  Version 2.64

[Important updates]
- [COLOR=#0000ff][B]Mitigate risk of security vulnerability related to [/B][/COLOR][COLOR=#ff0000][B]DRAM Row Hammering[/B][/COLOR][COLOR=#0000ff][B].[/B][/COLOR]

...

Just found it interesting that there now are BIOS updates that mentions fixes in relation to the security aspect of row hammer. I did try some testing before and after. But the 1 hammer error I get is so intermittent that I did not get any good data.

[RexFury117]

Must admit I never tested the new Kingston HyperX Savage sticks individually. So the 0 errors on hammer test 13 were with both sticks installed. Never increased the voltage though, and actually clocked them lower than the specs: @1600MHz 9-9-9-27-2T/1.5V (rated for @1866MHz).

If you are getting 0 errors when 1.5V. You unfortunately need to evaluate yourself what value the extra speed/timings with 1.65V, vs the hammer errors. It's tricky. If it was any of the other memory test than hammer, I personally would never accept any errors. When it comes to hammer, I will try very hard to never get it on new systems. But on old, like my laptop, I am now living with 1 single hammer error. Most stable laptop I have owned.

PassMark now has a good section about Hammer Test 13 in their Troubleshooting Memory Errors. Talks about single vs multiple RAM modules at the end of it. Worth a read.

Ok. I was asking because when i tested my current memory i got 0 hammer test errors when i tested each stick individually. However with both sticks installed i would get varying amounts of hammer errors depending on the ram voltage. So if u aren't getting errors with both sticks installed then that is an improvement over my memory. I put an order in for the 1866mHz hyperX savage memory.

[sveinan]

When you tested the hyperx savage ram did u get 0 errors when u tested the sticks individually or with both sticks installed?

Must admit I never tested the new Kingston HyperX Savage sticks individually. So the 0 errors on hammer test 13 were with both sticks installed. Never increased the voltage though, and actually clocked them lower than the specs: @1600MHz 9-9-9-27-2T/1.5V (rated for @1866MHz).

If you are getting 0 errors when 1.5V. You unfortunately need to evaluate yourself what value the extra speed/timings with 1.65V, vs the hammer errors. It's tricky. If it was any of the other memory test than hammer, I personally would never accept any errors. When it comes to hammer, I will try very hard to never get it on new systems. But on old, like my laptop, I am now living with 1 single hammer error. Most stable laptop I have owned.

PassMark now has a good section about Hammer Test 13 in their Troubleshooting Memory Errors. Talks about single vs multiple RAM modules at the end of it. Worth a read.

[RexFury117]

@sveinan
And the Kingston HyperX Savage 2400 16 GB kit passed as well. Good thinking on grabbing something newly introduced!

When you tested the hyperx savage ram did u get 0 errors when u tested the sticks individually or with both sticks installed?

[RexFury117]

I just recently upgraded my system and picked up some Crucial Ballistix Elite 16GB (8GBx2) RAM, http://www.microcenter.com/product/404938/Ballistix_Elite_16GB_DDR3-1866_(PC3-14900)_CL9_Desktop_Memory_Kit_(Two_8GB_Memory_Modu les). After increasing the RAM voltage to 1.65v in my BIOS, the best result I can get is around 20 errors during the Hammer Test. 0 errors with all other tests. If I only have 1 stick in though, I will get 0 errors during the Hammer Test. All errors experienced are all single bit errors.

[renzore101]

Do you mind giving the model # and retailer? I just bought some G.Skill Ripjaws to replace G.Skill Snipers and they both fail miserably. I think Ill be sending the Ripjaws back for credit and purchasing what you bought. I have a similar setup (Asus mobo w/ 4790k).

Hello folks, after many hours spent troubleshooting, hoping various Windows file corruption issues I was experiencing were not hardware related, I finally decided to run memtest. Sure enough with G.Skill Ripjaws I got over 10,000 bit errors. Today I purchased a pair of Crucial Ballistix Sport XT 1866 sticks and 1 hour into memtest I have a fraction of the amount of errors. Currently I have 39 errors that look like they exclusively occured in the "hammer" test as well. I am going to continue to monitor this but I am beginning to think G.Skill is not a good bet for future memory purchases. Passmark, great product!

[sveinan]

And the Kingston HyperX Savage 2400 16 GB kit passed as well.

Nice One more data point

[Armchair]

@sveinan
And the Kingston HyperX Savage 2400 16 GB kit passed as well. Good thinking on grabbing something newly introduced!

[Armchair]

A little late. But what you ordered is same series as mine, just binned even better probably. You wouldn't want my retailer/price here in Norway anyway For reference, the series I bought from was Kingston HyperX Savage. Basic idea: Newly introduced 'high-end' series, newly produced, newer chips, better tested for row hammer (maybe). I went for the DDR3-1866 HX318C9SRK2/16 2x8GB (PDF spec). I just want to run as close to 'standard' specs as possible (stability/quality then performance). One of the XMP profiles on the DDR3-1866 sticks is DDR3-1600 9-9-9-27-2T/1.5V, using that.

As you probably have observed you have bought the same DDR3-2400 HX324C11SRK2/16 2x8GB (PDF spec) as user klmd a few posts back (he got 0 errors on them). Not many datapoints to brag about though. But I hope and think you'll be ok. Hopefully this series will prove to be tested/resistant to row hammer going forward. For those that care.


Yeah, I saw that. Very creative usage :/ Since last time, someone has also created a good Row hammer Wikipedia page (MemTest86 6.0.0 mentioned at bottom of it). A little disappointed not more about it in tech press.

Yeah, Norway would have taken a bit of shipping time I wanted to get the order in before the weekend. I noticed Klmd had ordered the DDR-2400 as well and it was cheaper then the 1866 at Amazon so I thought why not.