Processors

[thekevinguy1]

I recently upgraded from a 3600 to 5800x and now have experienced occasional crash and restarts. When it happens, the colors look like the 'deep fried' memes (colors dark, inverted) before restarting like normal. People having similar issues attributed it to psu undercapacity, but whereas they had a 650w I have a 850w which should sufficient going off multiple calculators. In addition:

• I have run stress tests for both cpu and gpu with no crashes over 10 minutes of testing.
• Bios and chipset up to date
• This has only happened while playing several games.
• At load temps are 65-70C with a noctua d15s
• It always crashes with the graphical issues and this has never happened with the 3600

Any help would be appreciated.

[ThreeDee]

Always list your full system specs when asking for help with your PC .. BIOS version .. etc

What exact make/model 850w PSU are you using?

Are you plugged into a UPS?

What GPU and IF it requires more than one power connector, are you running separate power cables from your PSU to each power input on your GPU?

How is your air flow?

Are you using any cable extenders INSIDE your case and/or a PCI riser cable?

Are all your power inputs on your motherboard populated? (CPU power at top of motherboard)

Is your Windows install up to date?

I always go into device manager and under View , check "Show hidden devices" .. under Processors .. I'll remove all instances of my old processor (might help, might not )

[thekevinguy1]

Thanks for responding! My specs and answers are below:

Specs

• 5800x
• Noctua NH-D15s
• TUF GAMING X570-PLUS (WI-FI) (Latest Bios 4408)
• 4x8GB Corsair Vengeance 3600
• EVGA XC3 GeForce RTX 3080 10GB
• corsair rm850x 850W 80+ gold
• Fractal Meshify C
• 2 x 140mm Fans, 1x 120mm Fan
• Windows 10

Responses

1. PSU corsair rm850x
2. No UPS
3. Separate GPU power cables
4. Air flow is adequate
5. No extenders or risers
6. All power populated
7. Windows 10 but up to date
8. Devices uninstalled

[ThreeDee]

Does your issue happen with DOCP disabled/off? 

Does your  issue happen IF you up the voltage on your RAM from from 1.35v to 1.36v..1.37v..etc with DOCP enabled?

Does your issue happen with just 2 sticks of RAM installed in slots A2/B2? (2nd and 4th slots away from the CPU socket)

Have you tried reseating your CPU? (check for bent pins too)

Are you running ASUS's Armory crate software? .. if so, have you tried uninstalling it and see if issue persists?

[thekevinguy1]

Thanks for getting back to me again

1. Have not tried turning of DOCP
2. Did not increase RAM voltage
3. Have not tried 2 ram slots

However, the issue happens slightly less than once per day, so it is hard to verify that it is fixed besides seeing that it isn't fixed. Additionally, if either of these memory changes fix the problem, does that speak to the root of the problem or would I just be sacrificing performance.

Thanks!

[ThreeDee]

IF issue go away with DOCP turned off .. than that could point to the IMC on your new chip

Bumping up voltage on your RAM might rectify things to make DOCP stable

manually setting FCLK to 1800 might help IF it's not auto setting your FCLK correctly to match your RAM speed for some reason

Turning off DOCP and messing with RAM voltage and checking/setting FCLK are quick easy BIOS settings you can adjust with out having to do anything but a reboot

[thekevinguy1]

I'll do some testing when I have time and thanks again for the help. So far no crashes in the last day but I have no reliable way to trigger or reproduce it. If I continue to have any issues I'll try additional steps.

Thanks

[thekevinguy1]

Thanks for the response and my specs and answers are below:

Specs:

• 5800x
• Noctua NH-D15S
• TUF GAMING X570-PLUS (WI-FI) (BIOS 4408)
• 4x8GB Corsair Vengeance 3600
• EVGA XC3 GeForce RTX 3080 10GB
• corsair rm850x 850w 80+ Gold
• Fractal Meshify C
• 2 x 140 Fans, 1 x 120 Fan

Responses

1. Separate power cables
2. Air flow is adequate
3. No cable extenders
4. All power inputs populated
5. Windows 10 but up to date

[thekevinguy1]

Adding a reply in case anyone has this same issue.

I haven't had the issue in a week or so after making several changes, but will update if it returns. Primarily, performing a memtest with DOCP auto enables produced several memory errors which other users reported as well as a potential cause besides faulty power supply. I didn't try replacing the PSU out of convenience, but besides a defect it really should have been within its capabilities. 

To attempt to resolve the memory issues, I manually set the speed and timings, and increased voltage all the way to 1.4V, which finally led to passes in memtest. Hopefully that solves the problem and this error no longer happens.

[elstaci]

I have the same Motherboard as you do except the "Pro" version.

I needed to manually overclock my RAM to its native 3600Mhz speed by the drop down menu in BIOS Settings. I just manually set the actual speed to 3600Mhz and FCLK to 1800Mhz without any issues.

But once after manually setting the RAM I decided to use DOCP in BIOS which I located. Big mistake. As soon as I did my PC would reboot and my RAM would reset to its SPD speed. I then manually overclocked my RAM and again no more issues.

What was interesting after updating to my current BIOS Version 4021 I couldn't find DOCP anymore. Unless it was relocated to some obscure setting that I generally wouldn't look in.

At the moment I have BIOS version 4021 installed. I have downloaded the latest but I am waiting to see if there are many reports of it being buggy before upgrading.

NOTE: When you upgraded to the 5000 series CPU did you look to see if your RAM was listed in your motherboard's QVL list for the 5000 series processors or at the RAM Manufacturer QVL list if it wasn't list on the Motherboard QVL list?

Could be a reason for your unstable PC due to RAM not being 100% compatible with the new processor.