Showing results for 
Search instead for 
Did you mean: 

Read this before flashing BIOS 3701 - X299

Level 7
Good morning.

Decided to upgrade the BIOS on my X299 TUF Mark-1 from 3601 to 3701 because it was claiming better RAM compatibility and this happened...

Boot Looping to BIOS Setup - basically "dead"

1. BIOS upgrade "failed" alarming short 8-beeps (allegedly VRAM issue);
2. RAM Slot DIMM_C1 not being recognized;
3. BIOS version 3701 in fact way worse in CPU stability/performance;

1. Removed all RAM sticks and kept only 1 on DIMM_A1 slot;
2. Cleared CMOS;
3. Flashed BIOS 3701 again;
4. Pressed "OK_RAM!" Button to force RAM training;
5. Placed a 2nd RAM stick on DIMM_C1 (paired with DIMM_A1 according MB manual);
6. Trained RAM again with "OK_RAM!" button;
7. Boot to Windows and it was "fine";
8. Placed all 4 RAM sticks and "OK_RAM!" button again;
9. Entered Windows 11 UEFI normally;
10. Configured BIOS and RAM Timings as it was before this debacle;
11. CPU started to crash during gaming;
12. Some games reported not being able to access VRAM in DX12;
13. Downgraded BIOS back to 3601;
14. Trained RAM again;
15. Reset BIOS Setup and RAM Timings to what I was using before;
16. After a couple Windows 11 updates .NET and Cumulative from Jul/2022 PC now is back running stable at 4.3GHz/1.100V.

Final Thoughts:
It seems the BIOS 3701 messed with ReBAR and CPU microcode became an issue together with UEFI hardware direct access.
So, I would strongly recommend skipping this BIOS version 3701 altogether.

Edit 1: Microcode seems to have been changed also which made my CPU 10C warmer at same clock 4.5GHz (regardless BIOS version).

Edit 2: System no longer stable at 4.5GHz... will try to flash another BIOS aiming bring back previous microcode... wish me luck!

Edit 3:
7820X 4.5GHz (max. stable clock)
CPU Volt. 1.125V (voltage for above clock)
Celsius 83C (temperature under load)

7820X 4.3GHz (max. stable clock)
CPU Volt 1.100V (voltage for above clock)
Celsius 93C (temperature under load)
RAM timings and voltage were kept the same...
Crashes on DX12 - no idea how to fix this - any help would be more than welcome

Level 7
After a week or so, I found out the root cause for all this... my CPU cooler (Cooler Master Airmaker 8 - 1.5kg cooler) is too heavy and was pressing in excess on the processor socket.

I was able to restore it back to the original operational status after fully disassembling the PC and putting it back together with minimum torque on the mounting screws.


Level 9
Hi. First of all, are sure that you have mark 1 mobo, maybe mark 2? And secondly, are you saying that the problem was the cooler, but not updating your bios to the bios 3701, so it was just a coincidence?