Extreme + 1101 + 32GB G.Skill F5-6000J3636F16GX2-TZ5RK

[Serza]

My memory kit XMP I profile already has this set to 96.
I just ran another quick test with VDD and VDDQ set to 1.375 - Prime failed after 1 hour and 16 minutes.
I have now set VDDQ TX Voltage to 1.375.
If this doesn't work, I might try and loosen the timings from 36 to 40.

[Serza]

Update ...

Asus Lighting Service was hogging some CPU and the lights were moving in slow motion again.
Killed the lightning service manually using task manager.
Re-started crate and seemed back to normal.
Selected re-start PC in Windows 11 (soft reboot)
Then ...
Stuck on “Detect Memory: 01”
All resets from case button would do nothing but pass the CPU tests then get stuck again on Detect Memory: 01
Forced power-of using hold power button for 4 seconds method.
Re-started: all back to normal again.

Set XMP I
Asus MCE: Off
Changed VDD and VDDQ to 1.375
Re-start Prime 95 Test - Large FFTs no AVX
Side Note: According to hwinfo64, the temp range for this memory kit is 0C - 85C.
Prime95 failed at 1 hour 16 minutes.
Temps go to 82C for memory and 69C for CPU and 75C for CPU Package.
The overall system however was still working normally.
Software re-start – no problems.

I then un-installed ALL armoury crate software
Update IVR Transmitter Voltage: 1.375
Note: No longer have any lighting on Case strip, Pump and CPU Block.
Prime 95 Large FFTs No AVX; no errors after 2 hours 30 minutes.
Temps reached: 79C for memory and 69C for CPU and 75 for CPU Package.

I then stopping Prime95 to run a series of game benchmarks.
All settings max everything.
Res: 3840x1600 144Hz G-Sync Monitor

FC5 Benchmark (min/avg/max)
5 back-to-back runs with FPS average 120
GPU: 65C / CPU: 60C

FC New Dawn Benchmark (min/avg/max)
5 back-to-back runs with FPS average 113
GPU: 62C / CPU: 55C

FC6 Benchmark (min/avg/max)
5 back-to-back runs with FPS average 94
GPU: 64C / CPU: 52C

AC Odyssey Benchmark (min/avg/max)
5 back-to-back runs with FPS average 73
GPU: 65C / CPU: 53C
Note: ACOD did not properly exit - needed to task manager kill process
Note: HWInfo64 also stopped - need to be restarted

AC Valhalla Benchmark (min/avg/max)
5 back-to-back runs with FPS average 69
GPU: 64C / CPU: 50C

Memory temp average across all tests @ 52C
These tests took a little over 1 hour to setup and run.
--- Software Reboot, no problems
Currently running without problems.

So in summary, settings are;
XMP I
Asus MCE – Off
VDD @ 1.375
VDDQ @ 1.375
Update IVR Transmitter Voltage: 1.375
OS Changes: Not running any Asus Crate software

[Silent Scone@ROG]

Update ...

Asus Lighting Service was hogging some CPU and the lights were moving in slow motion again.
Killed the lightning service manually using task manager.
Re-started crate and seemed back to normal.
Selected re-start PC in Windows 11 (soft reboot)
Then ...
Stuck on “Detect Memory: 01�
All resets from case button would do nothing but pass the CPU tests then get stuck again on Detect Memory: 01
Forced power-of using hold power button for 4 seconds method.
Re-started: all back to normal again.

Set XMP I
Asus MCE: Off
Changed VDD and VDDQ to 1.375
Re-start Prime 95 Test - Large FFTs no AVX
Side Note: According to hwinfo64, the temp range for this memory kit is 0C - 85C.
Prime95 failed at 1 hour 16 minutes.
Temps go to 82C for memory and 69C for CPU and 75C for CPU Package.
The overall system however was still working normally.
Software re-start – no problems.

I then un-installed ALL armoury crate software
Update IVR Transmitter Voltage: 1.375
Note: No longer have any lighting on Case strip, Pump and CPU Block.
Prime 95 Large FFTs No AVX; no errors after 2 hours 30 minutes.
Temps reached: 79C for memory and 69C for CPU and 75 for CPU Package.

I then stopping Prime95 to run a series of game benchmarks.
All settings max everything.
Res: 3840x1600 144Hz G-Sync Monitor

FC5 Benchmark (min/avg/max)
5 back-to-back runs with FPS average 120
GPU: 65C / CPU: 60C

FC New Dawn Benchmark (min/avg/max)
5 back-to-back runs with FPS average 113
GPU: 62C / CPU: 55C

FC6 Benchmark (min/avg/max)
5 back-to-back runs with FPS average 94
GPU: 64C / CPU: 52C

AC Odyssey Benchmark (min/avg/max)
5 back-to-back runs with FPS average 73
GPU: 65C / CPU: 53C
Note: ACOD did not properly exit - needed to task manager kill process
Note: HWInfo64 also stopped - need to be restarted

AC Valhalla Benchmark (min/avg/max)
5 back-to-back runs with FPS average 69
GPU: 64C / CPU: 50C

Memory temp average across all tests @ 52C
These tests took a little over 1 hour to setup and run.
--- Software Reboot, no problems
Currently running without problems.

So in summary, settings are;
XMP I
Asus MCE – Off
VDD @ 1.375
VDDQ @ 1.375
Update IVR Transmitter Voltage: 1.375
OS Changes: Not running any Asus Crate software

You may want to look at getting some airflow over the DIMMs, however, if stable for your use case you can safely ignore Prime.

[Serza]

Yeah case setup kinda blocks air flow to the DIMMs - The 3 fans on the bottom feed directly into the 3080 (standard mount - not vertical).
The 3 on the side rad hit the res. and I am using the DIMM.2 for NVMe. The DIMM.2 module is large and kida blocks flow into the DIMMs.
The 3 on the top rad exhaust.
The rear fan is exhaust - perhaps I could make that intake ? (does not have a dust filter but those are cheap as to get ...).
I could also make it one of those Noctua industrial fans (if I could tolerate the noise) ...
DIMMs don't get over 55 during normal use (games etc.)
I do remember seeing a post by corsair when DDR5 was just starting up and they said that yeah DDR5 runs hot...

[Silent Scone@ROG]

Yeah case setup kinda blocks air flow to the DIMMs - The 3 fans on the bottom feed directly into the 3080 (standard mount - not vertical).
The 3 on the side rad hit the res. and I am using the DIMM.2 for NVMe. The DIMM.2 module is large and kida blocks flow into the DIMMs.
The 3 on the top rad exhaust.
The rear fan is exhaust - perhaps I could make that intake ? (does not have a dust filter but those are cheap as to get ...).
I could also make it one of those Noctua industrial fans (if I could tolerate the noise) ...
DIMMs don't get over 55 during normal use (games etc.)
I do remember seeing a post by corsair when DDR5 was just starting up and they said that yeah DDR5 runs hot...

The PMIC being onboard the module adds an extra layer of complexity when overclocking, but it depends on your setup and how conditional the overclock is. It never hurts to keep things cooler, though.

[Nikos4Life]

Good afternoon,

I have the same equipment configuration and it is impossible for me to make the XMP profile work stable with the RAM running at 6000 MHz.

12900K
Asus Extreme 1101 BIOS
F5-6000U3636E16GX2-TZ5RS (white version)

I have read the values you have suggested through this thread but still no way.

I have tried the following values:

XMP I
Asus MCE – On
VCCSA From 1.1 to 1.25
VDD @ From 1.30 to 1.40
VDDQ @ From 1.30 to 1.40
Update IVR Transmitter Voltage: From 1.35 to 1.40
Memory controller Voltage: From 1.25 to 1.45
Mem Tweaks I&II

No way to pass MemTest


Anything else I can try before thinking my CPU is the reason?


PS: MemTest with Auto values (no XMP) passes perfectly fine.



Thanks in advance

[gimbal]

May also want to try the built-in Memory Presets for Samsung (located under DRAM Timing Control).

[Nikos4Life]

May also want to try the built-in Memory Presets for Samsung (located under DRAM Timing Control).

Going to try it out, Should I use the profile with Manual or XMP and then apply the profile?

[gimbal]

Since the objective is really to get XMP I running and stable, I would enable XMP I, in addition to one of the Samsung presets. There are actually 2 presets for the 6000 memory so you could try both. Not sure what the difference is between the two.

[gimbal]

It also depends on what frequency and timings the memory presets set it to. Not sure if they set it to 6000/36 or not. So the other option is to set the presets just by themselves and see what happens.