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.

529 Views
1,738 REPLIES 1,738

Tsikos wrote:
Yesterday i build my new rig with asus rog x670e-e strix board and i have problem with network, for some reason the ethernet auto disable the internet and must reboot the system to get internet again but this happened every 10-20 minutes after boot any solutions?


I had this same issue. Go to device manager and open the 10G NIC click advanced tab and disable both Recv Segment Coalescing ipv4 and ipv6.

Tyberious wrote:
I had this same issue. Go to device manager and open the 10G NIC click advanced tab and disable both Recv Segment Coalescing ipv4 and ipv6.


I have the Intel Ethernet Controller I225-V controller and cant see this option in advanced tab, maybe have a diferent name?

Tsikos wrote:
I have the Intel Ethernet Controller I225-V controller and cant see this option in advanced tab, maybe have a different name?


Ah, sorry I thought you had the Marvell 10g. Sorry, I am sure it's a similar issue, but I don't know the options on the intel chips. Someone from Asus will need to help with that one since my quick research says it's a top reported issue with that chip.

Tyberious wrote:
Ah, sorry I thought you had the Marvell 10g. Sorry, I am sure it's a similar issue, but I don't know the options on the intel chips. Someone from Asus will need to help with that one since my quick research says it's a top reported issue with that chip.


No problem, but also my wifi stop work after some time when the ethernet lost the conection!!!
Asus please help us!!!

Nico67
Level 7
I have been trying to get per core boost limiter work with an Async103.6+mlb tune, however I can't get it to limit anything like I have set it. usually it doesn't limit at all, or it ends up dropping from 55.5x down to 55x giving me 5698, or lastly 4599 which doesn't seem to tie up with anything. I have tried setting it to straight frequency like 5675 or 5675/1.036, but it doesn't seem to limit a core as I would have expected. I started with 0801 bios but tried 0041 and it gives the same result.
I am wondering if it doesn't work with external clk enabled, may try that next just to see it work.

Nico67 wrote:
I have been trying to get per core boost limiter work with an Async103.6+mlb tune, however I can't get it to limit anything like I have set it. usually it doesn't limit at all, or it ends up dropping from 55.5x down to 55x giving me 5698, or lastly 4599 which doesn't seem to tie up with anything. I have tried setting it to straight frequency like 5675 or 5675/1.036, but it doesn't seem to limit a core as I would have expected. I started with 0801 bios but tried 0041 and it gives the same result.
I am wondering if it doesn't work with external clk enabled, may try that next just to see it work.


i presume u are not oc mode (not setting ratios manually) if u override ratios it wont work
it doesnt take into account bclk, did you do the appropriate adjustments?
u pin the workload to a specific core?

Shamino wrote:
i presume u are not oc mode (not setting ratios manually) if u override ratios it wont work
it doesnt take into account bclk, did you do the appropriate adjustments?
u pin the workload to a specific core?


Async 103.6 + mlb and 5439 DOS
Max should be 5749 ish and I tried setting things like 5675 or 5478 (less the blk) and run CB23 single with affinity locked to a core.

Hows does it implement the limit? does it fix to the nearest .25 ratio like 54.75?

Nico67 wrote:
Async 103.6 + mlb and 5439 DOS
Max should be 5749 ish and I tried setting things like 5675 or 5478 (less the blk) and run CB23 single with affinity locked to a core.

Hows does it implement the limit? does it fix to the nearest .25 ratio like 54.75?


yes now i see the same thing
when the bclks are different in asynch mode, when they are different then i couldnt find a freq iinput that wouldnt over shoot or undershoot alot.
i will have to check if it is possible to co exist per core limit with asynch bclk.

Shamino wrote:
yes now i see the same thing
when the bclks are different in asynch mode, when they are different then i couldnt find a freq iinput that wouldnt over shoot or undershoot alot.
i will have to check if it is possible to co exist per core limit with asynch bclk.


Had a chance to play with straight PBO and it does seem to work by limiting the requested core frequency, I did find it dropped the effective frequency way down, but I believe I probably need to reduce to the CO offset to compensate.

IE: PBO+200 all cores boost to 5750, but some don't run that all the time under load, requesting 5725 down to 5700. If I reduce one down to 5675 the effective freq was 5525ish, but only using 1.395vcore, so stretching heavily. I think I would need to reduce the -27 CO offset to something like -24, but still setting up CO's first to see what I can run.

Would be nice to get a extra 50-100mhz on the pbo offset, +200 is limiting the best cores a touch 🙂

Shamino wrote:
yes now i see the same thing
when the bclks are different in asynch mode, when they are different then i couldnt find a freq iinput that wouldnt over shoot or undershoot alot.
i will have to check if it is possible to co exist per core limit with asynch bclk.


0803 seems alot better, Async 104.1 5777.5max with alot of cores set to 5725, just under 5725.5 bin seems to hold pretty well, but not entirely 🙂
Its crazy how much more negative CO offset you can then give them to make them all run pretty much 5710-5725 effectual running CB23 single affinity set to test. However they do still have issues with very light loads and HWINFO does still see occasional full request max spikes to 5777.5, don't think the cores try it though.

Tried voltage suspension static floor set to 1.3v, but it does seem to work with power saving features of PBO and is over ruled by DOS settings at 1.29v, even though I have C-States disabled. Might try high performance power plan to se if that has any effect.