I take back what I said about 1304 bios

[wing_man]

I'm back to what I had before 1304 update. XMP 2 set at 5800mhz. It seemed stable yesterday at 6000mhz but today I'm back to 5800mhz to keep it stable.

[wing_man]

wing_man wrote:
I'm back to what I had before 1304 update. XMP 2 set at 5800mhz. It seemed stable yesterday at 6000mhz but today I'm back to 5800mhz to keep it stable.

I found out what happened yesterday. I had made the three changes in Extreme Tweaker\Dram Timing control\Memory Training Algorithms that Buildzoid had suggested. I had xmp 2 set at 6000mhz. I rebooted and tested the memory and everything was fine. I never shutdown the pc. I found the next time I started the pc it was back to 5800mhz for stability.

This help stability if you never shutdown the pc
SenseAmp offset Training Disabled
Round Trip Latency Enabled
Turn Around Timing Training Disabled

[wing_man]

wing_man wrote:
I'm back to what I had before 1304 update. XMP 2 set at 5800mhz. It seemed stable yesterday at 6000mhz but today I'm back to 5800mhz to keep it stable.

I had to up VDD & VDDQ to 1.4v and now I can run 6000mhz with XMP 2.

Apex build date 11-21
12700k UEFI 1304
Samsung CL36 36 36 76 1.3v

[Silent_Scone]

Hello,

Same kit @ 1.45v VDD/VDDQ 1034. Karhu Ramtest stable

[rolloff]

Silent Scone@ROG wrote:
Hello,

Same kit @ 1.45v VDD/VDDQ 1034. Karhu Ramtest stable

And a few restarts later, the same setting can suddenly become unstable...
Btw: If you have read the Overlock.net forum carefully, you now know that there is a problem with the z690 Apex Boards that were manufactured before 2022. The boards with a later production achieve significantly higher and more stable RAM frequencies. So posting any screens here without information from the board revision is pointless.

With a board of almost $800, I feel downright screwed. Not only does the RAM run significantly better on the new revisions, you also have to deactivate any energy-saving mechanisms for the PCH in order not to be spammed with WHEAs.

The fact that ASUS does not allow an RMA in this case means that ASUS have lost a long-standing customer with me.

[Silent_Scone]

rolloff wrote:
And a few restarts later, the same setting can suddenly become unstable...
Btw: If you have read the Overlock.net forum carefully, you now know that there is a problem with the z690 Apex Boards that were manufactured before 2022. The boards with a later production achieve significantly higher and more stable RAM frequencies. So posting any screens here without information from the board revision is pointless.

With a board of almost $800, I feel downright screwed. Not only does the RAM run significantly better on the new revisions, you also have to deactivate any energy-saving mechanisms for the PCH in order not to be spammed with WHEAs.

The fact that ASUS does not allow an RMA in this case means that ASUS have lost a long-standing customer with me.

Sounds like you're projecting your own stability problems there. If settings "become" unstable then the overclock is too conditional. Potentially training drift/temperature induced. The same timing spacing denoted above and voltages may not work for all kits or CPUs.


6800 stable on same kit. This is a day one release board. All C-States enabled also. Again, this sounds like your overclock is unstable. Instability with resuming from S3 is often memory instability related. Trying to lay blame on any one vendor with philosophical opinions when it comes to overclocking is difficult as each case is different.

[Falkentyne]

Silent Scone@ROG wrote:
Sounds like you're projecting your own stability problems there. If settings "become" unstable then the overclock is too conditional. Potentially training drift/temperature induced. The same timing spacing denoted above and voltages may not work for all kits or CPUs.


6800 stable on same kit. This is a day one release board. All C-States enabled also. Again, this sounds like your overclock is unstable. Instability with resuming from S3 is often memory instability related. Trying to lay blame on any one vendor with philosophical opinions when it comes to overclocking is a waste of ones time, some users tend to do this on any subject which isn't easily understood.

There was some discussion about this.
Apparently a fix to this is, when changing memory or even CPU overclock settings, to save, go back to BIOS, turn off the system and then completely remove the PSU power for 30 seconds, then flip the PSU back on and turn on the system. Two people seem to have tested this with positive results.

Also another user said that an Asus rep they contacted confirmed that "a problem exists" with certain board runs (not the Hero capacitor thing), and accepted a RMA, but didn't give any details.

[VroiteR]

wing_man wrote:
I'm back to what I had before 1304 update. XMP 2 set at 5800mhz. It seemed stable yesterday at 6000mhz but today I'm back to 5800mhz to keep it stable.

you downgrade ur bios? if yes, which one? and wow u did that u have a guide. i never did a downgrade. ty

[wing_man]

VroiteR wrote:
you downgrade ur bios? if yes, which one? and wow u did that u have a guide. i never did a downgrade. ty

No, I just set frequency back to 5800. I now have it set to 6000 with voltage increase.

[Silent_Scone]

Not sure what part of the post you're referring to.

Instability between changing settings is not uncommon. Occasionally data will be stuck in NVRAM that can be cleared by AC cycles. It's unlikely this behaviour occurs when the system is stock, too (because there is no unstable OC).

I can't speak on behalf of regional reps, so anything else you would need to contact service for your region or post in the correct section for service.