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.

567,696 Views
2,373 REPLIES 2,373


@bwana wrote:

@SkyBeam thanks for your explanation. But is it possible to know which modules ASUS validated those timings in the bios for? For example, they list SK hynix 2x16 at various speeds. Obviously they tested specific modules. If I have the same modules, could I then not just use that setting?


These are not MODULES. In fact these are timings for certain memory CHIPS as of my knowledge. There is a difference between modules and chips in a sense that a module might limit the timings or clock rates due to layout, noise handling etc. So the chip timings are just there in case you know exactly which chips are on your module and you might want to try. There is no warranty at all that those chip timings are working with your module even if your module is equipped with the exact same chips.

In addition most people don't know which chips are on their modules as most of them are covered by heatsinks. You might get your chip vendor information from SPD-EEPROM information in BIOS or tools like CPU-Z but still you won't know exactly which chips are on your modules. Some 3rd party modules might even mix and match chips from differend batches depending what is available on the market. So it's impossible for these profiles to work on just all modules, even if you get a DDR5-5600 module and know you have SK-Hynix chips you might not be able to run the profile for SK-Hynix DDR5-5600 2x16GB...

 


, what do you use for memory testing? Karhu?

Don't know Karhu but googling it I would not want to use it. Any tool running under Windows will have inherent limits to test the complete memory as some memory is used by Windows, Applications and Kernel. So you cannot test those areas by just writing/reading from it. Also the security model of modern operating systems prevent an application to just read random from any area (like kernel or other application area) for good reason.

I do recommend to just get a free copy of memtest86+ and create a boot USB-Stick and boot it directly. Memtest86+ will load into GPU memory area in order to be able to test the complete RAM. Yes, granted some iGPU might also cover some parts of the RAM. But at least a much smaller part than a fully blown OS is reserving.


@Tsikos wrote:

Why asus can't fix all that i have this system 7 months and still unstable, i paid 1500 euro to upgrade from x570 to x670e and i think i made a big mistake!




RMA. If you are having problems with your motherboard, RMA it. 

Don't feel bad. A lot of people are RMAing these motherboards. It's really the only way that Asus knows that something is wrong with their product. 

Is it normal to go from perfectly stable settings to no-POST the next day like that? I'm running 2x32GB (F5-6000J3040G32GX2-TZ5RK) on ROG CROSSHAIR X670E EXTREME with 7950X. I'm running with DOCP II, Context Restore + Power Down and 105W power limit on CPU (AMD Eco mode). Everything seems ok (restarts, power off even cutting power to PC entirely), but when I let it sit for longer period powered off and boot straight into memtest86+ I get memory errors in first few seconds/minute (say 20 errors) or no-POST at all (sometimes it takes just reboot, sometimes I have to clear CMOS).

JEDEC memory speeds seem OK, but I certainly did not need to spend that much on board and RAM if I that was all I wanted.

This happened to me even with my x670 Hero and 7950x , although the only change i made to bios was using EXPOII , i then compared latest 1004 bios defaults to the previous always stable 0805 amd only change i noticed imedeatly was thar CPU VSOC was set the same as the RAM that is 1.35v as oposed to before on 0805 was 1.25v for CPU VSOC / 1.35v MEM v , so i set manually this CPU VSOC to 1.25v let all other values as they were as defaults on 1004 bios and now its perfectly stable.

 
Conclusion: at least in my x670 HERO with Ryzen 7950x i cant have  1.35v on CPU VSOC same a RAM , causes PC to shutdown totally like a short circuit on some loads like games heavy on CPU usage. 

Right, I forgot about that. I did set VSOC to 1.25V after upgrading to 1002 together with MEMCLK DIV to 1:1.

kamyk155
Level 10

Today first for a long time I had problem with computer start. Long beep and three shotrt beeps - GPU problem.

Weird - the computer started without restart and any info about it - just beeped and go forward to windows.

Bios is 1004 on my strix e-e. 

aphysically
Level 8

where do you find the "overclocking presets" for X3D in the asus bios; I'm having trouble finding it. I have x670e proart creator The thing in the OP mentioned as for "X3D OC Preset" for "simple slight perf boost for X3D"

N7Valiant
Level 7

Might just be my setup, but when I enable the X3D Core Flex gaming presets my OS will freeze up without fail 30-60 seconds after boot.  I enabled pretty much every other OC setting other than AI Overclocking or anything that requires manual settings.

Changing Algorithm 1 from Enabled => Auto fixes the freezing issue.

This is on Arch Linux, so it might be a niche issue.

CPU is AMD 7900X3D.

Were you able to find the overclocking presets mentioned in the post? Where is it in the BIOS settings?

Deluxef
Level 7

So I did series of experiments to better understand the "cold boot" (no-POST or memory errors after prolonged power off of the machine). The hardware is once again: ROG CROSSHAIR X670E EXTREME, R9 7950X with G.SKILL F5-6000J3040G32GX2-TZ5RK. The BIOS settings were the same for all of the experiments: DOCP I, with VSOC set to 1.25V, MCR+PowerDown enabled, UCLK=MCLK and 105W AMD Eco Mode.

ZenTimings_Screenshot_28016007.4945552.png

The theory is that there is some temperature related issue with the memory, which causes instability at *LOW* temperatures (think room temp). To test this I shut down the machine, take the RAM out, wait for 10 minutes, put everything back together (taking note to put each RAM module in the exactly same slot - by their SN) and boot straight into memtestx86+.

- First, I let the machine sit powered off and nothing else -> POST and memtest with no issues.
- Repeat the same thing, but taking the memory out and putting it back -> POST and memtest with no issues.
- This time I took the RAM out and put it in the fridge (to cool it quickly before the rest of the machine) -> no-POST (with 0xFE Q-code) or memtest errors for minute or so after the boot.

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.