cancel
Showing results for 
Search instead for 
Did you mean: 

X670 resource

Shamino
Moderator

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:
110000
instead of 0011000
after selection press downcore apply changes or discard if made mistake

ocpak/octools

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


Test BIOSes:

new:
X3D OC Preset for those MB with asynch BCLK Support: (for simple slight perf boost for X3D)
97792

DOCP/EXPO Tweaked: (for simple timings tightening)
97793

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:

dualseg.jpg

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

97403

97404

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.

1,012 Views
2,382 REPLIES 2,382


@Deluxef wrote:

I've repeated the experiment with taking the RAM out with and without cooling it down twice. Every time I did *NOT* cool the RAM I got normal boot without memory errors. With the cold RAM I got no-POST or at least memory errors. So it seems to me like there is some memory temperature compensation issue on ASUS or G.SKILL side.


ASUS might sell memory pre-heater units soon 🙂

Joke aside this might be the reason why it works for some by not using MCR as the training itself is heating up the sticks already. However ASUS might be able to solve the problem also by some temperature-related curve adaptations. Perhaps just adding 0.05V more to the memory VDD/VDDQ might help avoiding this cold-boot issue.

Semiconductors usually get more conductive when warm, so they need slightly more voltage when cold and voltage should drop when they heat up. This is also a reason for thermal runaway if driven by constant voltage (temperature going up --> resistance going down -> more current -> temperature going up -> resistance going down --> more current...)

Its pretty weird, as I can go lower with VDD/VDDQ without issue (1.35V, while the memory kit is 1.4V). However at 1.4V there is no much headroom to go up (BIOS limit seems to be at 1.4375V). I've also tried to loosen a bit on timings (like 30-40-40 -> 32-40-40) and frequency (6000 -> 5600), but still this "cold boot" issue seems to persist. I may have to pick up another memory kit and hope its not bios/board/cpu issue (then again, I haven't seen any other 2x32GB 6000 with 30-X-X timing).


@Deluxef wrote:

Its pretty weird, as I can go lower with VDD/VDDQ without issue (1.35V, while the memory kit is 1.4V). However at 1.4V there is no much headroom to go up (BIOS limit seems to be at 1.4375V). I've also tried to loosen a bit on timings (like 30-40-40 -> 32-40-40) and frequency (6000 -> 5600), but still this "cold boot" issue seems to persist. I may have to pick up another memory kit and hope its not bios/board/cpu issue (then again, I haven't seen any other 2x32GB 6000 with 30-X-X timing).


VDD/VDDQ lower voltages is not a surprise to me. Remember that 1.35V is already pretty much overvolted. JEDEC specification for DDR5 is 1.1V. My G.Skill DDR5-6000 CL30 also use 1.35V by default and I can also run at 1.30 in most cases but starting to see some memory errors in memtest86+ occasionally - which do not lead to immediate crashes; so some people might believe it's "fully strable" while it isn't.

For AMD systems the most important memory-related parameter is the memory controller voltage (MC voltage). Try to play with this one at DDR5-5600 first. Note: You should not load one of the optimized timing profiles as they will be very tight on timings typically. So just select DOCP/EXPO proile 1 and then go for DDR5-5600 and potentially lower timings slightly and increase MC voltage to 1.400V. Or even try 1.45V. Your 32GB modules put quite some more load on the memory controller as the more common 16GB modules. But granted my 4x16GB configuration will put even more load on the MC and it's working fine with 1.400V here. But I guess it also comes down a bit to the silicon lottery.

Also some people reported their SOC voltage to be too high (>1.35V) and lowering it might have stabilized their system. This is why I am running in offset mode at about -0.075V.

pz123456
Level 10

i was so frustrated, i returned my x670 hero and bought a x670 extreme.

the issues are still the same, even with bios 1004

-after loading the bios, no screen signal while loading windows ( you see the circle which is loading windows, then screen goes off and you have to hard reset the pc (reset button, sometimes turn off power supply) 
-when trying CoreFlex and asynchronous BCLK with 101 instead of 100, AIDA Stresstest crashes immediately (no curve optimiser, DOCP I Ram 6000 C36 GSkill QVL.
I tries running everything on default - but my blank screen issues persists. I even tried the GPU with another mainboard (x570 Extreme) and i NEVER had this problem with blank screen. 

Does anybody know a solution?? im tired of hard resettting my pc, when screen goes blank when entering windows. 

when will the next beta bioses come? i think ill return this board too and buy a gigabyte or msi

I have similar issues, no matter what bios I use the get a 00 code on a reboot.

Have to hard reset to keep going.

Asus support is the worst. The made me PAY for an advance rma, gave me shipping info then cancelled it!!!!


@pz123456 wrote:

i was so frustrated, i returned my x670 hero and bought a x670 extreme.


So you were so frustrated to give back the board and you bought the very same thing again?
Not sure I will ever understand this. Actually the boards just differ in some features like brackets, buttons and slots but essentially it's thee same layout and for sure running on exactly the same (BIOS) codebase. So not sure what one would expect swapping it out for the same thing.

If you don't like it you should actually buy something completely different - e.g. swapping it out for an Intel platform or move to another vendor.

 


@pz123456 wrote:

-after loading the bios, no screen signal while loading windows ( you see the circle which is loading windows, then screen goes off and you have to hard reset the pc (reset button, sometimes turn off power supply) 
-when trying CoreFlex and asynchronous BCLK with 101 instead of 100, AIDA Stresstest crashes immediately (no curve optimiser, DOCP I Ram 6000 C36 GSkill QVL.
I tries running everything on default - but my blank screen issues persists. I even tried the GPU with another mainboard (x570 Extreme) and i NEVER had this problem with blank screen. 

Does anybody know a solution?? im tired of hard resettting my pc, when screen goes blank when entering windows.


Actually your description is contradictionary.  You saay your screen signal goas away while loading windows and in the same sentence you write you see the loading circle. Seeing the loading circle on the primary GPU main screen is expected behavior. No signal would be if your screen entering standby mode and there is actually NO video signal.

Make sure to select your primary GPU in BIOS. Remember that you will always have 2 GPUs installed (iGPU and dGPU) if you install any GPU in your PCIe slots. Also remember that you will see the BIOIS and Windows loading circle only on the main screen on primary GPU (depending on whether you selected iGPU or PCIe GPU to be primary in BIOS).

If your screen goes black only AFTER having seen the loading circle - so actually when Windows has loaded the GPU drivers and enters high-resolution mode - then your issue might be a completely different one. You might just run on a screen configuration where Windows shows output only on screens which are not connected or the screens are not on this input. I have seen people connecting HDMI + DP to the same screen. Then Windows was configured to use the DP screen as primary and HDMI as secondary (Windows thinking there are 2 screens connected) . But actually the screen was configured to the HDMI input and was showing the Windows second screen only (which is pitch black on Windows 11 logon screen). So users believed it's "broken" or off. But actually you could move your mouse around and see it appear on the screen. It was just not the primary one.

I would recommend to flip screen configurations. If Windows loads try using Win+P buttons to toggle screen configurations. I typically recommend to hold down Win and then Press "P" twice. Then let go the Win key and wait a few seconds for WIndows to re-arrange screens. If you eventually hit the "Clone" option then all your screens should show the same output. If this worked, then go to your Windows Screen configuration and switch to "Extend" mode arranging your screens the way you want and also selecting which one is supposed to be your main/primary.

Another thing you should check is your cabling. You might try first with a single screen. Connect it via ONE cable to your GPU, no matter if iGPU or PCIe graphics card. But use a single cable to a single screen only and make sure the screen is configured to use the input you are connecting it to. Windows should then (in any case) select this single screen as the main one. If Windows believes there is any other screen it might use this one to extend your desktop to.

Another caveat might be that more and more screens come with USB-C connectors. USB-C however is not only carrying USB signals. It can also carry a DisplayPort video signal. So if you hook up your screen to your PC using HDMI AND also USB-C then Windows might think there are 2 Screens connected. It might use the USB-C/DP-alt-mode screen as primary so you won't see anything on your HDMI output as it is used as second screen only. That's why I am saying: Try first connecting your screen only with a SINGLE cable to your PC first (despite power cord of course).

Tank you for your response - im sorry -  i didnt explain it very well. 


 



Actually your description is contradictionary.  You saay your screen signal goas away while loading windows and in the same sentence you write you see the loading circle. Seeing the loading circle on the primary GPU main screen is expected behavior. No signal would be if your screen entering standby mode and there is actually NO video signal.

Make sure to select your primary GPU in BIOS. Remember that you will always have 2 GPUs installed (iGPU and dGPU) if you install any GPU in your PCIe slots. Also remember that you will see the BIOIS and Windows loading circle only on the main screen on primary GPU (depending on whether you selected iGPU or PCIe GPU to be primary in BIOS).

 



I see the BIOS Screen and after that  the circle loading windows. Then my Screen Turns off (thats what i meant with blackscreen). My 4k Screen is connected with one DP Cable (no second Monitor, no HDMI). I tried unplugging the cable and using HDMI - its the same result. You can hear that windows is loading (start sound) but i cant get a monitor signal at this point. I disabled the IGPU on my previous hero - i had the same issues with my blackscreen (monitor turning off). 

and i know that crosshair and hero are nearly the same boards - but hero has no dual bios. in the past i spend hours of installing new bios versions. I have the x570 extreme - which i never had problems with my gpu (4090 Tuf). 

greetings

//Edit: Win+P when screen turns off is not working. I have to reset the pc


@pz123456 wrote:

//Edit: Win+P when screen turns off is not working. I have to reset the pc


Did you try what happens when you press the ATX power-button briefly? Will Windows eventually shut down properly after a few seconds and your PC turn off? If it does, this will be the proof that it actually booted and working just fine. Just there is no video output or more precisely there might be video output which cannot be displayed by your screen. So you need to debug your screen configuration, not your mainbaord.

If nothing happens when your press the ATX power button and/or also you can't toggle your numlock light on the keyboard, then your PC might be freezing up/crashing and you might indeed look at some hardware/mainboard related issues - or memory instabilities. Try running memtest86+ for a few hours to check for basic stability.


@pz123456 wrote:

I see the BIOS Screen and after that  the circle loading windows. Then my Screen Turns off (thats what i meant with blackscreen). My 4k Screen is connected with one DP Cable (no second Monitor, no HDMI). I tried unplugging the cable and using HDMI - its the same result. You can hear that windows is loading (start sound) but i cant get a monitor signal at this point. I disabled the IGPU on my previous hero - i had the same issues with my blackscreen (monitor turning off). 


That actually means you do have an output during boot and also connected the screen to the correct DP - as you wouldn't see the circling Windows loading screen.

As of your description you seem to lose signal just when the driver loads. There could be multiple reasons for this:

  • Your GPU driver does not work
  • Your GPU will use a resolution/refresh-rate combination which is not supported by your monitor or cable

One difference between AMD and nVidia drivers is that AMD drivers by default use the recommended refresh rate and resolution of a monitor while nVidia defaults to 60Hz usually.

So you say your screen goes off right after the loading circle. Do you see your screen showing any notification before going off like "refresh rate out of range" or similar? If yes then your screen might just report refresh rates which are not supported by itself or the cable. Even though you claim to have changed the cable it might be possible that both of your cables are not suitable. Do you use the cable included with the monitor or another one? If not then I would also recommend to get a decent DisplayPort 1.4 certified cable.

Another thing you could try is to disable Adaptive Sync in your monitor configuration if you have such an option enabled.

If this does not help, then try to boot your windows in recovery/safe-mode (usually it enters recovery automatically if you interrupt the boot process 3 times). Try DDU to remove any driver and then try to run with the default Microsoft Generic Display Adapter first and then try to install latest AMD adrenaline drivers (download from AMD website, not from Windows Update).

It is of course possible you might look at some sort of hardware fault but it's very unlikely to be a mainboard fault at this point as your display output seems to work and also Windows seems to load properly. I really think you are looking at some screen/refresh-rate/resolution misconfiguration or potentially cable issue. Both of your cables (DP/HDMI) might be unsuitable for the resolutions selected and there is no way for your PC to detect that the signal it puts on the cable might not get to your screen properly. So I recommend using a short and high quality cable and disable gaming features in your monitor temporary. If you are on a high-res screen (e.g. 4K@144) you might consider temporary using another screen too. For example connect a second 1080p60 screen to HDMI just to see it's getting a signal there. If yes, then configure extend mode and if your 4K screen goes black again, then lower the refresh rate of the 4K screen until you get an image.

So what I am saying is that your screen might tell your GPU that it can handle 4K@144 but actually there might be reasons this will not work (e.g. cable). None of this is mainboard related.