G.Skill F5-7600J3646G16GX2-TZ5RK XMP not stable ROG Z790 Extreme UEFI 0813

kill_a_wat · Feb 14, 2023

I would try and experiment with different voltages eg. Lower TX and IMC volts like what QXE mentioned and see if that helps. 7600Mhz should be very doable on Z790 Hero and Extreme.

scracy · Feb 14, 2023

kill_a_wat said:
I would try and experiment with different voltages eg. Lower TX and IMC volts like what QXE mentioned and see if that helps. 7600Mhz should be very doable on Z790 Hero and Extreme.

Already tried that prior to starting this thread, I may have to wait until a new UEFI is released for my board, interestingly according to G.Skill QVL for my RAM only for Apex no other ROG/Asus boards yet this memory still appears on Asus Z790 Extreme QVL

Anyway many thanks to all of you that have tried to help

QXE · Feb 14, 2023

scracy said:
Already tried that prior to starting this thread, I may have to wait until a new UEFI is released for my board, interestingly according to G.Skill QVL for my RAM only for Apex no other ROG/Asus boards yet this memory still appears on Asus Z790 Extreme QVL Anyway many thanks to all of you that have tried to help

you would be better off manually tuning. XMP is almost never stable.

logikju · Feb 15, 2023

Have you tried to increase the vcc SA a little bit and check again when running memtest?
Mid 60 doesn't seems too hot.

affxct · Feb 15, 2023

logikju said:
Have you tried to increase the vcc SA a little bit and check again when running memtest?
Mid 60 doesn't seems too hot.

💀

logikju · Feb 15, 2023

affxct said:
💀

I don't see any risks pushing it to 1.30v.

I mean there must be an explanation or at least a workaround as a lot of hynix A-die sticks reviews were done on the same Z790 Extreme board and could be pushed up to 7800 so it might be related while loading an xmp profile beyond 7400. Maybe setting it up manually as suggested by QXE.

The QVL is not exhaustive, I am running with this F5-6000J3038F16GX2-TZ5NR amd expo which neither appears on the gskill or gigabyte compatibility list and runs at stock or tweaked loaded expo profile on an z790 elite ax board.

affxct · Feb 15, 2023

logikju said:
I don't see any risks pushing it to 1.30v.

I mean there must be an explanation or at least a workaround as a lot of hynix A-die sticks reviews were done on the same Z790 Extreme board and could be pushed up to 7800 so it might be related while loading an xmp profile beyond 7400. Maybe setting it up manually as suggested by QXE.

The QVL is not exhaustive, I am running with this F5-6000J3038F16GX2-TZ5NR amd expo which neither appears on the gskill or gigabyte compatibility list and runs at stock or tweaked loaded expo profile on an z790 elite ax board.

Unless I’m crazy, 1.3 is actually the auto value for mid-7000 data rates when trained 😬. The issue with your statement is regarding temps. Mid-60s is utterly atrocious. There’s no shot anyone is running 7600 1usmus/VST stable with mid-60s. I mean maybe with XMP terrible subs it’s possible, but how is he in the 60s to begin with?

QXE · Feb 15, 2023

scracy said:
Gave it a go with a fan on the memory, temps much better but I still have an error, looks like I'm stuck with 7400Mhz XMP Tweaked
View attachment 2600054

Run 1usmus v3, also Test 1 = IVR TX afaik

logikju · Feb 16, 2023

affxct said:
Unless I’m crazy, 1.3 is actually the auto value for mid-7000 data rates when trained 😬. The issue with your statement is regarding temps. Mid-60s is utterly atrocious. There’s no shot anyone is running 7600 1usmus/VST stable with mid-60s. I mean maybe with XMP terrible subs it’s possible, but how is he in the 60s to begin with?

What I am saying is that the error didn't occurred because of the temp itself. Even with a fan didn't help.

Most of the screenshots shows a max temp of 39°c on both sticks with the same error. I just don't understand why suddenly on one screenshot the temps shows one stick running 11°c hotter than the other one.

scracy · Feb 16, 2023

affxct said:
Unless I’m crazy, 1.3 is actually the auto value for mid-7000 data rates when trained 😬. The issue with your statement is regarding temps. Mid-60s is utterly atrocious. There’s no shot anyone is running 7600 1usmus/VST stable with mid-60s. I mean maybe with XMP terrible subs it’s possible, but how is he in the 60s to begin with?

There is not a lot of airflow in my case as the P.C is water cooled combine that with 28 degrees C or more ambient temperature as it is summer here in Australia that why temps at worst was 64 C from memory. According to G.Skill anywhere from 40 to 60 degrees C is considered normal operating temperature for DDR5. The error is still there with lower temps so I dont think its temperature related.....not great temps but not much I can really do about it short of putting a fan there permanently. The screenshot with higher max temp on one stick would likely be because its the stick nearest the DIMM.2 so gets heat from the NVME drives and less airflow from the fans infront of the radiator.

affxct · Feb 16, 2023

logikju said:
What I am saying is that the error didn't occurred because of the temp itself. Even with a fan didn't help.

Most of the screenshots shows a max temp of 39°c on both sticks with the same error. I just don't understand why suddenly on one screenshot the temps shows one stick running 11°c hotter than the other one.

Ohhhhh nvm that's a bug. Ok no forget what I said. I thought they were actually legit in the 60s. That's just a common bug that everyone experiences.

affxct · Feb 16, 2023

scracy said:
There is not a lot of airflow in my case as the P.C is water cooled combine that with 28 degrees C or more ambient temperature as it is summer here in Australia that why temps at worst was 64 C from memory. According to G.Skill anywhere from 40 to 60 degrees C is considered normal operating temperature for DDR5. The error is still there with lower temps so I dont think its temperature related.....not great temps but not much I can really do about it short of putting a fan there permanently. The screenshot with higher max temp on one stick would likely be because its the stick nearest the DIMM.2 so gets heat from the NVME drives and less airflow from the fans infront of the radiator.

The 63.8c thing is actually just a bug. I realize that now. With regards to the normal temp thing, I'm not concerned about IC health or anything like that, but if you were in the 60s-70s it would be a lot harder to stabilise 7600 in hard tests.

kill_a_wat · Feb 16, 2023

AFAIK 63.8c temp spike is a bug. Doubtful you can stabilise much at all if avg DIMM temp was 60c+

logikju · Feb 16, 2023

scracy said:
There is not a lot of airflow in my case as the P.C is water cooled combine that with 28 degrees C or more ambient temperature as it is summer here in Australia that why temps at worst was 64 C from memory. According to G.Skill anywhere from 40 to 60 degrees C is considered normal operating temperature for DDR5. The error is still there with lower temps so I dont think its temperature related.....not great temps but not much I can really do about it short of putting a fan there permanently. The screenshot with higher max temp on one stick would likely be because its the stick nearest the DIMM.2 so gets heat from the NVME drives and less airflow from the fans infront of the radiator.

As the wc experts call it push pull, would it help adding one or two fans on your rad in order to increase airflow?

scracy · Feb 16, 2023

logikju said:
As the wc experts call it push pull, would it help adding one or two fans on your rad in order to increase airflow?

Potentially I could do that or water cool the memory, looking into it atm.

G.Skill F5-7600J3646G16GX2-TZ5RK XMP not stable ROG Z790 Extreme UEFI 0813

kill_a_wat

scracy

QXE

logikju

affxct

logikju

affxct

QXE

logikju

scracy

affxct

affxct

kill_a_wat

logikju

scracy

Top Contributors this Month

Recommended Communities