7800X3D with Gigabyte X670 AORUS ELITE AX GREEN SCREEN Crashes without ERROR messages.

[Shadowized]

Kernel-Power event ID 41 Task category 63.

yeah those are expected, seems fine. Guess you'll just need to give it a little time to see if it's all good.

 

[xastunts]

Happy now?

View attachment 367306

your timings isn't stable for me. get game crash, but not system crash.

 

[xastunts]

Hi everyone! I like to say that I have solved the problem with the crashes, and it was all because of a "too high" GPU MEM Clock. I have swapped out the main PSU and tested with another PSU and now swapped back to the main PSU again. No crashes on the replaced PSU, and no crashes on the original PSU as well.
I am so happy to finally have nailed down the culprit, and I know I couldn't have done it without all of you.
Thank you so much to all of you that posted.

What I need to do know is just running some stress test and perhaps getting a better tighter ram timings.
Case closed!

 

[xastunts]

Happy now?

View attachment 367306

Hi @TheLostSwede
Since you and I have the same or at least similar ram kit, I'm now sharing you my very tight timings for the Team Force Delta RGB 6000 model with stock XMP CAS Latency: 30-36-36-76 at 1.35v
I have spent a lot of time fine tuning the Team force delta 6000 KIT to achieve these ultra low timings!
the ram voltage vdd and vddq can not be as low as on my photo because I found getting random reboots when voltage were this low. I can indeed pass a full 25 cycles of 1 usmus v3 TM5 check without any errors. You may try 1.50v or 1.55v. I had random reboot at 1.45v, so the sweet spot could be somewhere between 1.46- 1.55v
and I would set the vddp to 1.05v, and mem vddio set to auto. my VSOC is 1.25v.
I have tried but couldn't pass a TM5 TEST at 6400 FCLK 2133, but I've been told it's possible with my kit.
You could try a tRFC at 356 like me and see if you got a really good bin or not. If that doesn't work, you can try higher until it posts, or until you pass a TM5 run of 25 cycles.

For all of you wondering about the high voltage for these timings, let me please explain.

To be able to achieve this very tight timings we will need to give the RAM more voltage or it won't even post. And to have it become stable we need to probably give it nothing below 1.5v for the VDD and VDDQ.

Now this is all in the safe zone for this particular ram as the temp is very low during these voltages. I've tried it at max 1.65v that my Gigabyte Aorus X670 Elite AX board can take for the RAM. With 1.65v I had my temps not exceeding 50 when during a stress test with a fan on in a hot ambient room temp of 30 degrees.
operating temps should be under 40 while now on a stress test.

you always want to monitor your temps if you are into over clocking and applying more voltage to the CPU or RAM.

I hope you like these timings, and I would like to know if you can run this on your bin, as each bin are different and maybe not able to copy ones timings.

Link to my ram at NewEgg
Team T-Force Delta RGB 32GB (2 x 16GB) 288-Pin PC RAM DDR5 6000 (PC5 48000) Desktop Memory Model FF4D532G6000HC30DC01 - Newegg.com

 

[Space Lynx]

I have spent a lot of time fine tuning the Teamforce delta 6000 KIT to these ultra low timings!
the ram voltage vdd and vddq can not be as low as on my photo as I could indeed pass full 25 cycles of 1 usmus v3 TM5 check without any errors. You may try 1.50v or 1.55v. I had random reboot at 1.45v
and I would set the vddp to 1.05v mem vddio set to auto. my VSOC is 1.25v.
I have tried but couldnt pass a test at 6400 FCLK 2133, but I've been told its possible.
You could try a tRFC at 356 like me and see if you got a really good bin or not.if that doesnt work you can try higher until it posts or until you pass a tm5 run.

you really shouldn't be recommending 1.5 volts for ram to people. i mean even at stock 1.35v my 6000 cl 30 ram hits 45-48 celsius when gaming, maybe I am remembering wrong, but isn't ram supposed to never pass 50 celsius otherwise stability issues may occur? @ir_cow

 

[xastunts]

you really shouldn't be recommending 1.5 volts for ram to people. i mean even at stock 1.35v my 6000 cl 30 ram hits 45-48 celsius when gaming, maybe I am remembering wrong, but isn't ram supposed to never pass 50 celsius otherwise stability issues may occur? @ir_cow

Hi! I'm sorry first of all. I think I should have been more specific when I replied to @TheLostSwede . I am now editing my original post and hope that further miscommunications can be avoided.

 

[ir_cow]

you really shouldn't be recommending 1.5 volts for ram to people. i mean even at stock 1.35v my 6000 cl 30 ram hits 45-48 celsius when gaming, maybe I am remembering wrong, but isn't ram supposed to never pass 50 celsius otherwise stability issues may occur? @ir_cow

JEDEC allows up to 85c at 1.1v. Could there be Long-Term degradation going above that voltage? Sure, but other factors like the IC brand, if it has cooling and at what temperature will determine the lifespan of the part.

If we are ignoring the lifespan of a part based on voltage, The next thing is some memory chips don't like more voltage like Micron. if you put 1.4 into that it will just throw errors. So really right now the only IC that accepts 1.4+ is SK Hynix. More voltage means more heat to disapate. The "limit" is still 85C, but some of the timings like CAS, tTRC and tREFI if not following the JEDEC spec will error out well before 85c.

I'm often having problems above 45c when overclocking for Ambient cooling, say it needs 1.65v to run. The second it hits 45~, errors appear. Either could drop the frequency or raise the CAS. Now at 1.65v it doesn't error until 60~ (or whatever). So it's a give and take. Need voltage for raw frequency, but your more likely to encounter errors if the voltage is higher due to other timings.

 

[Space Lynx]

JEDEC allows up to 85c at 1.1v. Could their be Long-Term deration going above that voltage? Sure, but other factors like the IC brand, if it has cooling and at warm temperature will determine the lifespan of the part.

If we are ignoring the lifespan of a part based on voltage, The next thing is some memory chips don't like more voltage like Micron. if you put 1.4 into that it will just throw errors. So really right now the only IC that accepts 1.4+ is SK Hynix. More voltage means more heat to disapate. The "limit" is still 85C, but some of the timings like CAS, tTRC and tREFI if not following the JEDEC spec will error out well before 85c.

I'm often having problems above 45c when overclocking for Ambient cooling, say it needs 1.65v to run. The second it hits 45~, errors appear. Either could drop the frequency or raise the CAS. Now at 1.65v it doesn't error until 60~ (or whatever). So it's a give and take. Need voltage for raw frequency, but your more likely to encounter errors if the voltage is higher due to other timings.

makes more sense now, thanks

 

[eidairaman1]

Hi everyone! I like to say that I have solved the problem with the crashes, and it was all because of a "too high" GPU MEM Clock. I have swapped out the main PSU and tested with another PSU and now swapped back to the main PSU again. No crashes on the replaced PSU, and no crashes on the original PSU as well.
I am so happy to finally have nailed down the culprit, and I know I couldn't have done it without all of you.
Thank you so much to all of you that posted.

What I need to do know is just running some stress test and perhaps getting a better tighter ram timings.
Case closed!

You need to contact asus about the issue, and were you pushing the vram beyond the boost speeds? If not then you should request rma of asus

 

[Redwoodz]

You need to contact asus about the issue, and were you pushing the vram beyond the boost speeds? If not then you should request rma of asus

He had fast timings selected. Glad he got it sorted. Screen freezes usually point to gpu.

 

[eidairaman1]

He had fast timings selected. Glad he got it sorted. Screen freezes usually point to gpu.

It is in regard to fix he found, his card being asus.