Overclock.net › Forums › Intel › Intel CPUs › *Official* Intel DDR4 24/7 Memory Stability Thread
New Posts  All Forums:Forum Nav:

*Official* Intel DDR4 24/7 Memory Stability Thread - Page 7

post #61 of 5441
Feel free to start one, the info in the first post is not proprietary. However, there are only two of us on the forum that understand memory timings and we cannot be in all places at once or answer everything as people would like us to.
post #62 of 5441
Thread Starter 
Quote:
Originally Posted by cookiesowns View Post

Where's the one for X99 confused.gif


I didn't feel the thread would serve it's purpose as well if it was platform agnostic, it would have been possible to add in other platforms but then that could have created congestion with results and confusion. Maybe we can add in X99..., but I draw the line at DDR4 smile.gif
Edited by Silent Scone - 9/2/15 at 1:43am
post #63 of 5441
Hello

Based on the first post, quoted below, Skylake only would be keeping with the intent of the thread. The DDR4 architecture is the only common link between X99 and Skylake. There is no comparison value as far as what may be optimal regarding obtainable speeds, timings or performance across the two platforms.

Quote:
Originally Posted by Silent Scone View Post

This threads main purpose is to both discuss information and various results and to gauge what is possible between different configurations, DIMM capabilities and CPU samples. Results are welcome all the way up the frequency spectrum. If it's obtainable, it should be posted! smile.gif.
post #64 of 5441
Thread Starter 
Yes, even with the use of a separate sheet I do not think there would be much benefit for X99 at this stage in the platforms life given there are plenty of results for people to gauge what is reasonable. Keeping the thread proprietary means that all discussion is kept to Skylake which hopefully will make it easier for people to digest looking for information
post #65 of 5441
So I've had some interesting findings with trying to get XMP working on my DD4 3200 corsair set with a Asus Z170-A motherboard.

When I first built the PC I had to up the SA and IO voltage to get it to post with XMP to 1.25 SA and 1.20 IO.

After awhile I tried lowering the voltages after seeing someone killed their 6700k somehow and it seemed to be memory controller related.

It would be unstable with auto/lower voltage than 1.25SA and 1.20IO with the XMP profile enabled.

Though the system was stable at higher voltages, my USB sound DAC was dropping sound like crazy. Every three seconds in games the audio would cut out and distort, come back and cut out and distort again. If I switched to onboard realtek sound I would get loud popping and extreme distortion after about 30 minutes of gaming.

I found last night when I lowered the SA and IO voltages the system wasn't stable but the USB DAC sound issues were gone and played all sound without a hitch. Turned the voltage back up and the cutting out and distorted sound came right back.

I found a beta bios update for my board yesterday that I upgraded to and it seems to have fixed all the problems.

Z170-A BIOS 0901
1.Update Intel ME version to 1168
2.Enhance DDR4 compatibility

After the upgrade I was able to turn SA and IO voltage back to Auto and I passed 4 hours of realbench and 4 hours of HCI memtest flawlessly with XMP on. Before the bios update the system would crash with XMP on at auto voltage. Sound is perfect now too.

So if you're getting audio issues and have added voltage to SA or IO that could be your issue.
post #66 of 5441
Thread Starter 
Quote:
Originally Posted by sauced View Post

So I've had some interesting findings with trying to get XMP working on my DD4 3200 corsair set with a Asus Z170-A motherboard.

When I first built the PC I had to up the SA and IO voltage to get it to post with XMP to 1.25 SA and 1.20 IO.

After awhile I tried lowering the voltages after seeing someone killed their 6700k somehow and it seemed to be memory controller related.

The System Agent voltage should be fine within 1.3v. I'd be interested to see how someone managed to conclude the CPU failure was memory controller related, and how much voltage was being applied throughout the system. Can't really comment on the audio issues vicariously, seems odd
post #67 of 5441
Can anyone comment on how much better the Google Stress App is at detecting instability compared to HCI memtest? I have used the free version of HCI for years and found it much better than memtest86. I don't have anything to boot Linux from atm but I can go pick something up if this test is clearly superior.
post #68 of 5441
Quote:
Originally Posted by Derp View Post

Can anyone comment on how much better the Google Stress App is at detecting instability compared to HCI memtest? I have used the free version of HCI for years and found it much better than memtest86. I don't have anything to boot Linux from atm but I can go pick something up if this test is clearly superior.

Stressapp is more stressful than HCI for testing memory alone. HCI's load is a bit more spread to the cache, while Stressapp is more memory bus isolating. They both have their use, but if testing memory alone, then Stressapp is a quicker way of finding errors.
post #69 of 5441
Thread Starter 
Yep, around 1500% pass will take you substantially longer than running stressapp
post #70 of 5441
Maximus VIII Extreme

Raja--6700K @4.8/4.8---DDR4-3733-C18-19-47-1T----1.40v---SA Auto & IO Auto ---Stressapptest----2 Hours and 1000% HCI pass:

14C water temp.




New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › *Official* Intel DDR4 24/7 Memory Stability Thread