Showing results for 
Search instead for 
Did you mean: 

X670 resource


ill use this thread to collect some new test bioses for the boards, maybe also to explain some less understood options

to disable cores ccd go here and choose ccd xx bit map down core.
each ones stand for an enabled core
best to disable from the back, ie:
instead of 0011000
after selection press downcore apply changes or discard if made mistake


7950x not boosting pass 5.5G -> check that CStates is not disabled
Detailed Explanation on CState Boot Limiter

Test BIOSes:

X3D OC Preset for those MB with asynch BCLK Support: (for simple slight perf boost for X3D)

DOCP/EXPO Tweaked: (for simple timings tightening)

strixe-e 1515 

strixe-f 1515 

strix e a 1515 

crosshair hero 1515 

crosshair gene 1515 

crosshair extreme 1515 

creator 670 1515

creator b650 1515

strix 650E I

strix 670 itx



for crosshair and strix e-e:

explanation of segment2 Loadline:


customize a heterogenous loadline for a dual segment workload range.

example above shows loadline=L6 when current is in range of 0~40A, and Level4 when current is above 40A.




Adds for x3d

dynamic ccd priority switch with core flex, os / driver agnostic so win10 win11 ok



Algo as follows:
If condition reached and ccd0 specified, then check current mem/cache activity > threshold and hysteresis reached, if fulfilled then switch
If condition reached and ccd1 specified, then check current mem/cache activity <=threshold and hysteresis reached,, if fulfilled then switch
Default hysteresis =4

Can combine multiple algos for ccd priority so combinations are wide

works on non x3d too but of course senseless on it. detailed explanation here.

2,404 REPLIES 2,404

And interestingly: 1710 started with x8, that was x16 after waking up after being in standby mode for a few hours.

Somehow the negotiation of the link width is not working properly, the link speed is always correct (PCIE 4.0). The author of this post presumes that would more likely point to a configuration problem than for example a noise-related problem.

My question is: could a hardware fault: IO-die, connection between socket and CPU or between socket and slot, or a broken graphics adapter (or any hardware defect) cause these link width problems with a clearly different probability and magnitude for firmware versions 1602-1807?

Level 9

DId anyone already try new 2214 BIOS on their TUF models?

Yes, but "Aio pump, fan headers" it's not working perfectly with

Do not use "AIO Pump header" or your pump will run at the minimum speed be cautious. configuring for maximum or custom speed won't work. this issue occurs in bios 2204 and 2214, rollback to BIOS 2001 and headers works perfectly.

Aio pump signal only because it is powered by 4pin

Bios 2214-2204

  • Aio pump and fan headers do not work properly.
  • Dram 6600C30, 6400C28, and 2:1 7600 are impossible 
  • Using Hynix A-Die

✔️Bios 2001

  • AioPump and Fans headers works perfectly.
  • Dram 6600C30, 6400C28 are working
  • Dram 2:1 7600 hard but is possible.
  • Low Soc Voltage requirement.
  • Using Hynix A-Die


Oh, that doesn't sound too promising - thanks for your time to check it out.
Since I am using an AIO pump that I definitely need to PWM control, I will stay away for now.

Level 11

I have the strix 670E-E and just went from 1609 to 1807 and it looks like the aura Stealth setting on shutdown is still broken. It won't hold the Stealth setting and the main led lighting stays on after shutdown. When you go into bios again the setting has reverted to "on". Anyone else had this issue?

There is a setting in armory crate under devices - shutdown effect that overrides bios setting.

Good to know, I will see if I can find that, cheers

Okay, just checked, was already off in AC, it would seem that that does not Override the bios setting, thanks anyway. Any idea where we can log bios issues? Not sure if there is an official thread here for that or not?

Level 10

Maybe someone know, can we disable new mitigation fix? cause sql DBs are very slow...

@fractal-zombie wrote:
Maybe someone know, can we disable new mitigation fix? cause sql DBs are very slow...

Normally, any mitigation can be disabled at the OS level. For Linux, you just pass 'mitigations=off' in the kernel parameters, or, if you compile the kernel yourself, you can simply disable all the mitigations in menuconfig...

As for disabling at the UEFI level, I am not sure it is possible at all. But a new option appeared with the mitigated AGESA, in "AMD CBS", "CPU common options", at the bottom of the settings page you will find a new "CPU speculative store mode" option, with Auto/Balanced/More speculative/Less speculative settings. I use "More speculative" myself, but it did not seem to make any measurable difference in the benchmarks I ran compared to "Auto" or "Balanced" (I did not try "Less speculative" so far). But I'm not using a SQL database either, so... See by yourself ! 🙂