RAM Overclock Prime95 Blend Shuts Off

[flamingspartan3]

I have G.Skill 1333 MHz CL9 (4 x 2 GB) RAM and am attempting to overclock it. It runs fine at the stock 9-9-9-24 timings at 1400 MHz. When I attempt to push it to 1600 MHz, I have to raise timings to 10-10-10-24. With these timings and frequency, it passes Memtest86 at the original 1.5 V. However, if I run the Prime95 Blend, the PC shuts off after a few minutes when the RAM is set to 1600. There is no blue screen and the Prime95 log doesn't appear to work. Does anyone have any tips about how to deal with this issue? Thanks!

Full specs:
Asus P8Z68-V LE
Intel Core i7 3770S
AMD R9 390
G.Skill 8 GB 1333 MHz RAM
Samsung 850 Pro SSD
Seagate 1 TB HDD
Antec HCG 620W modular PSU

http://www.newegg.com/Product/Product.aspx?Item=N82E16820231425

 

[JackNaylorPE]

use memtest86+ for memory testing so s to eliminate all other considerations.

As far as what to expect recognize that 13333 represents the very low end and pushing the speed may require use of higher voltages or timings.

 

[flamingspartan3]

use memtest86+ for memory testing so s to eliminate all other considerations.

As far as what to expect recognize that 13333 represents the very low end and pushing the speed may require use of higher voltages or timings.

What I don't understand is that why the PC just shuts off and there are no blue screens. Normally, memory errors can at least be documented/observed. The other Prime95 tests don't lead to any issues, only the blend test does. Raising the voltage to 1.575 (the max for Ivy Bridge) doesn't have any effect. Even though 1333 is low, many people have been able to overclock that RAM to 1600 and higher.

 

[JackNaylorPE]

This is the 1st time I have answered a question whereby a user was using P95 to test for RAM errors. Since it was (for Sandy Bridge and older) primarily a CPU stress test, there's a probability that the RAM is not the problem. When testing, as the goal is to eliminate as many outside influence as possible, you're starting off w/ a handicap.

As always, Event Viewer should be the 1st place you look for a shut down cause.

Many people have, many people haven't. To shorten the trial and error workload, I normally visit the manufacturer's web site and look to see what timings and other secondary adjustments people used to successfully OC their modules.

 

[flamingspartan3]

This is the 1st time I have answered a question whereby a user was using P95 to test for RAM errors. Since it was (for Sandy Bridge and older) primarily a CPU stress test, there's a probability that the RAM is not the problem. When testing, as the goal is to eliminate as many outside influence as possible, you're starting off w/ a handicap.

As always, Event Viewer should be the 1st place you look for a shut down cause.

Many people have, many people haven't. To shorten the trial and error workload, I normally visit the manufacturer's web site and look to see what timings and other secondary adjustments people used to successfully OC their modules.

I looked at the Event Viewer but the only notes are that "the shutdown was unexpected" and "The system has rebooted without cleanly shutting down first". What is more troubling is that there are no system logs between the time the Prime95 test started and the time that the shutoff occurred. The logs resumed after the reboot. I'm fairly sure this is a RAM issue because the shutdowns do not occur at stock frequency or even when overclocked to 1400 MHz at stock timings. Also, my CPU temperatures don't go above 75 on the blend test so I doubt there is overheating involved.

 

[JackNaylorPE]

Without a Memtest86+ test, there's no additional advice I can offer. I will again leave you with:

1. Visit your RAM manufacturer's web site / forum to see what speed / timings other users may have had success with or advice given by TS staff,
2. Run Memtest86+

 

[Themastererr]

Timings are still too fast. Try -
11
11
11
35

 

[flamingspartan3]

Without a Memtest86+ test, there's no additional advice I can offer. I will again leave you with:

1. Visit your RAM manufacturer's web site / forum to see what speed / timings other users may have had success with or advice given by TS staff,
2. Run Memtest86+

I ran Memtest86+ and it passed the test with no errors.

 

[flamingspartan3]

Timings are still too fast. Try -
11
11
11
35

Aren't those slightly high for 1600 MHz? Especially the 35?

 

[JackNaylorPE]

I ran Memtest86+ and it passed the test with no errors.

Dunno if ya read the source page but a memtest86+ test is usually done for 8 - 12 hours. If it passes, them memory is not your problem.

Aren't those slightly high for 1600 MHz? Especially the 35?

Not sure how you are referencing high / low but those are slower timings

 

[flamingspartan3]

I ran Memtest86+ and it passed the test with no errors.

Dunno if ya read the source page but a memtest86+ test is usually done for 8 - 12 hours. If it passes, them memory is not your problem.

Aren't those slightly high for 1600 MHz? Especially the 35?

Not sure how you are referencing high / low but those are slower timings

I ran the test for 8 hours and found no errors with the program. Also, I tried playing games and editing video with no blue screens or random shutdowns either. I think Prime95 just is acting strangely on the blend test since the other tests show no issues.

I just though that 35 was too high (slow) of a timing for 1600 MHz and it's possible to achieve lower (faster) ones. Either way, since I found no issues, I think I will try to push my overclock further to 1800 MHz by raising timings.

 

[Themastererr]

I ran Memtest86+ and it passed the test with no errors.

Dunno if ya read the source page but a memtest86+ test is usually done for 8 - 12 hours. If it passes, them memory is not your problem.

Aren't those slightly high for 1600 MHz? Especially the 35?

Not sure how you are referencing high / low but those are slower timings

I ran the test for 8 hours and found no errors with the program. Also, I tried playing games and editing video with no blue screens or random shutdowns either. I think Prime95 just is acting strangely on the blend test since the other tests show no issues.

I just though that 35 was too high (slow) of a timing for 1600 MHz and it's possible to achieve lower (faster) ones. Either way, since I found no issues, I think I will try to push my overclock further to 1800 MHz by raising timings.

Yes, I just told you just to go a very slow one (35) to rule out if the timings were in fact the issue. If it solved it, then you could go faster from there.

 

[flamingspartan3]

I ran Memtest86+ and it passed the test with no errors.

Dunno if ya read the source page but a memtest86+ test is usually done for 8 - 12 hours. If it passes, them memory is not your problem.

Aren't those slightly high for 1600 MHz? Especially the 35?

Not sure how you are referencing high / low but those are slower timings

I ran the test for 8 hours and found no errors with the program. Also, I tried playing games and editing video with no blue screens or random shutdowns either. I think Prime95 just is acting strangely on the blend test since the other tests show no issues.

I just though that 35 was too high (slow) of a timing for 1600 MHz and it's possible to achieve lower (faster) ones. Either way, since I found no issues, I think I will try to push my overclock further to 1800 MHz by raising timings.

Yes, I just told you just to go a very slow one (35) to rule out if the timings were in fact the issue. If it solved it, then you could go faster from there.

Now I'm not sure if it is actually solved. I tested with a different memory stress program, MemTest, and my PC shuts itself off there as well. I even increased the timings at 1600 MHz to 11-11-11-28 and the shutdown still occurs. I'm not sure what the problem is...

At stock RAM frequency, it doesn't shut off. Even though I have overclocked to 1600 MHz, I have also tried raising both the timings and voltage. It still shuts down.

 

[Themastererr]

I ran Memtest86+ and it passed the test with no errors.

Dunno if ya read the source page but a memtest86+ test is usually done for 8 - 12 hours. If it passes, them memory is not your problem.

Aren't those slightly high for 1600 MHz? Especially the 35?

Not sure how you are referencing high / low but those are slower timings

I ran the test for 8 hours and found no errors with the program. Also, I tried playing games and editing video with no blue screens or random shutdowns either. I think Prime95 just is acting strangely on the blend test since the other tests show no issues.

I just though that 35 was too high (slow) of a timing for 1600 MHz and it's possible to achieve lower (faster) ones. Either way, since I found no issues, I think I will try to push my overclock further to 1800 MHz by raising timings.

Yes, I just told you just to go a very slow one (35) to rule out if the timings were in fact the issue. If it solved it, then you could go faster from there.

Now I'm not sure if it is actually solved. I tested with a different memory stress program, MemTest, and my PC shuts itself off there as well. I even increased the timings at 1600 MHz to 11-11-11-28 and the shutdown still occurs. I'm not sure what the problem is...

At stock RAM frequency, it doesn't shut off. Even though I have overclocked to 1600 MHz, I have also tried raising both the timings and voltage. It still shuts down.

That's wacked out. G skill is known for being able to overclock well too...Something's missing. Don't give up just yet and sorry I couldn't help more

 

[flamingspartan3]

I ran Memtest86+ and it passed the test with no errors.

Dunno if ya read the source page but a memtest86+ test is usually done for 8 - 12 hours. If it passes, them memory is not your problem.

Aren't those slightly high for 1600 MHz? Especially the 35?

Not sure how you are referencing high / low but those are slower timings

I ran the test for 8 hours and found no errors with the program. Also, I tried playing games and editing video with no blue screens or random shutdowns either. I think Prime95 just is acting strangely on the blend test since the other tests show no issues.

I just though that 35 was too high (slow) of a timing for 1600 MHz and it's possible to achieve lower (faster) ones. Either way, since I found no issues, I think I will try to push my overclock further to 1800 MHz by raising timings.

Yes, I just told you just to go a very slow one (35) to rule out if the timings were in fact the issue. If it solved it, then you could go faster from there.

Now I'm not sure if it is actually solved. I tested with a different memory stress program, MemTest, and my PC shuts itself off there as well. I even increased the timings at 1600 MHz to 11-11-11-28 and the shutdown still occurs. I'm not sure what the problem is...

At stock RAM frequency, it doesn't shut off. Even though I have overclocked to 1600 MHz, I have also tried raising both the timings and voltage. It still shuts down.

That's wacked out. G skill is known for being able to overclock well too...Something's missing. Don't give up just yet and sorry I couldn't help more

Your timings for 1600 MHz (11-11-11-35) worked much longer than usual but the PC still shut down in Memtest even if the voltage was raised to 1.550 V. I've tried 1400 MHz (9-9-9-24) and it's stable at 1.50 V. I'll see if I can push any more but I doubt it.

 

[gskill support]

Can you post pictures of memory settings and timings?

Do you have the latest BIOS?