[CJMitsuki]

I can tell you the bios settings even once I am home, currently away. RTC seems to have issues on windows 1809 for me, gives me an error about initialization which sucks. Can just copy & paste the bios profile settings when I'm back for you to try out My IMC kinda sucks so I decided to see if I could lower the timings instead of increase MHz, wouldn't budge even 1mhz past otherwise it's touch and go, seems very temperature dependent.

If its the same initialization error Im thinking of then I found it had to do with another program that had implemented another version of the WinRing driver and there were conflicts or RTC was trying to use that particular driver. I want to say TurboVcore caused this but It may have been another tool similar.

 

[nick name]

If its the same initialization error Im thinking of then I found it had to do with another program that had implemented another version of the WinRing driver and there were conflicts or RTC was trying to use that particular driver. I want to say TurboVcore caused this but It may have been another tool similar.

My instance of the issue didn't involve any of those listed programs. If that helps at all.

 

[BLUuuE]

I can tell you the bios settings even once I am home, currently away. RTC seems to have issues on windows 1809 for me, gives me an error about initialization which sucks. Can just copy & paste the bios profile settings when I'm back for you to try out My IMC kinda sucks so I decided to see if I could lower the timings instead of increase MHz, wouldn't budge even 1mhz past otherwise it's touch and go, seems very temperature dependent.

Try installing Open Hardware Monitor

 

[crakej]

So I've been testing 3703MHz all day and it's been very very promising. Of course I have to test overnight, but I think today's results warrant it.

Edit:

Welp it looks the settings I was running yesterday became unstable today. The only difference between yesterday and today being increased ambient temperature.

What voltage did you use for the ram here?

 

[nick name]

What voltage did you use for the ram here?

I think I was running 1.56V for 3703MHz. I've stopped testing it after it seemed to be much too sensitive to temps. Unless there was some other variable I am not accounting for.

 

[gupsterg]

Continuing on from post 5192, at 5hrs 47min test failed.

Spoiler

I tried a few different things after that, but adding a VCORE offset seems to have got me past 5hrs+ barrier. I believe the offset is aiding training at POST to stabilise this OC. As when I increase VCORE offset, load voltage for me never changes, only idle.

Spoiler

@gupsterg

...had my first working day, so holidays were over for me right now.

Sucessfully tested the test ssd setup.

The overdriventool has only been used to put the fan spinning up to 1300rpm.
She likes it cool. (Fury)

Maybe i'll make a gaming based installment of Win 10 1803, and a productive one.

Actually i think it could be a combination of a corrupt stick,followed by some issues on the OS Level, and those may based on the software i have installed on the main system ssd.
BUT thats a thing i've to check out.
If it isn't that i'll tell ya.

So thank you very much for your help, i hope i can move on now, get my stable to that point where it was before my issues started.

Cheers mate

Spoiler

Gesendet von meinem SM-G950F mit Tapatalk

NP , hope all is still well.

@Keith Myers and all linux users:

This looks like a neat project to get ASUS WMI sensors readouts in linux https://github.com/electrified/asus-wmi-sensors

Requires BIOS 1002 on the C7HWIFI. I finally have a reason to update! What's the most stable recent BIOS now?

It'll be great to report results here and also provide feedback to the author!

Thanks for posting, will be trying it ASAP.

1002 if want AGESA 1.0.0.2C, 1103 if you want AGESA 1.0.0.6.

 

[Wuest3nFuchs]

@gupsterg

Hi, how are you?

First off all i checked the typhoonburner and it told me that my dram controller supports up to 1466mhz =2933. I didn't made a karhu stresstest yet, but i plan to in the next days.
Also tryin mint with gsat.

To be true ,today i had 4 failed shutdowns in a row on my main ssd .

On the fresh installed OS SSD i hadn't this issue.
So i dunno where this come from.
I also can see no issue on ob the event history .

Last weekend i ordered a 3 usb3.0 usb stick with different sizes and even 2x usb2.0. So i hopefully can eliminate a faulty OS installation.

I hold thumbs up for your stability and benchmark results!
btw what is this RT you always Talk about ?

cheers mate

Gesendet von meinem SM-G950F mit Tapatalk

 

[gupsterg]

@gupsterg

Hi, how are you?

First off all i checked the typhoonburner and it told me that my dram controller supports up to 1466mhz =2933. I didn't made a karhu stresstest yet, but i plan to in the next days.
Also tryin mint with gsat.

To be true ,today i had 4 failed shutdowns in a row on my main ssd .

On the fresh installed OS SSD i hadn't this issue.
So i dunno where this come from.
I also can see no issue on ob the event history .

Last weekend i ordered a 3 usb3.0 usb stick with different sizes and even 2x usb2.0. So i hopefully can eliminate a faulty OS installation.

I hold thumbs up for your stability and benchmark results!
btw what is this RT you always Talk about ?

cheers mate

Gesendet von meinem SM-G950F mit Tapatalk

Hi, all good here .

Officially Ryzen 2xxx supports 2933MHz on 2 dimms of single rank/sided RAM, but you can OC . I also do run GSAT on Mint as as well. You may find CPU-X handy on Linux, I used it in my screenies on that OS, it's like CPU-Z; Keith recommended it.

As you have no issue on fresh installed OS can only assume something is up with OS on main SSD. Dunno what to suggest to know what is wrong with that install.

Yes this chip has been great . I plan to upgrade radiator, as I reckon the PBO OC with BCLK 103.8MHz has pushed it. RT is abbreviation for Karhu RAM Test.

All the best and hope to read of you gaining back OC :thumb: .

 

[Keith Myers]

Thank you . . . . thank you

@Keith Myers and all linux users:

This looks like a neat project to get ASUS WMI sensors readouts in linux https://github.com/electrified/asus-wmi-sensors

Requires BIOS 1002 on the C7HWIFI. I finally have a reason to update! What's the most stable recent BIOS now?

It'll be great to report results here and also provide feedback to the author!

Thank you very much minal for the post. I was not aware of this new kernel mode sensor driver. I have stuck with BIOS 0702 on my C7H boards because the later version aren't compatible with the latest kernels.

I will be giving this new driver a test after updating to the 1002 BIOS for the C7H. It will be great to get all of the WMI sensors available in Linux.:thumb:

[Edit] Tried to compile but hit errors. Logged a issue on the repository. Will wait for an reply about how to fix the errors.

[Edit 2] The developer said the driver only will compile on kernels > 4.17 because of changes in HWMON. Said he will be fixing it shortly to work on lesser kernels.

 

[Keith Myers]

Is it possible to return to BIOS 0702 after updating to BIOS 1002 and determining that 1002 isn't compatible with Linux.

 

[Wuest3nFuchs]

hey keith! i used bios flashback from 0509 to 1201 and back to 0509. So yes it's possible. Hope that answer helps you.

Gesendet von meinem SM-G950F mit Tapatalk

 

[Martin778]

Hm, I have PE3 on, DOCP standard, memory set to 3333MHz Stilt Fast @ 1.4V, everything else on auto but I still very rarely see that 4.3GHz PBO frequency, pretty much only very brief jumps in Ryzen Master or HWInfo64.
When gaming or even running Cinebench R15 single core, all 8 lock to 4.15GHz and stay there.

 

[CJMitsuki]

Hm, I have PE3 on, DOCP standard, memory set to 3333MHz Stilt Fast @ 1.4V, everything else on auto but I still very rarely see that 4.3GHz PBO frequency, pretty much only very brief jumps in Ryzen Master or HWInfo64.
When gaming or even running Cinebench R15 single core, all 8 lock to 4.15GHz and stay there.

Thats because PE doesn’t rely on memory speed to determine its multiplier. Temps are what it uses. Aside from that PE3 won’t hit 43.5 all core unless you bclk OC to around 104mhz then to stay at 3333 you have to step memory down 2 steps. 1 step for each 2 MHz on the base clock to keep it at the same freq. PE3 is more for single core clocks and PE4 is for the higher all core clocks.

 

[Martin778]

What's weird is that when running, say 3dmark Port Royal, the 1st test shows the CPU mostly at 2.2GHz and only sometimes jumping to 4.1-4.35.

 

[CJMitsuki]

What's weird is that when running, say 3dmark Port Royal, the 1st test shows the CPU mostly at 2.2GHz and only sometimes jumping to 4.1-4.35.

That’s bc it’s barely using the CPU and the GPU is taking most of the workload unless on the physics test.

 

[nick name]

Hm, I have PE3 on, DOCP standard, memory set to 3333MHz Stilt Fast @ 1.4V, everything else on auto but I still very rarely see that 4.3GHz PBO frequency, pretty much only very brief jumps in Ryzen Master or HWInfo64.
When gaming or even running Cinebench R15 single core, all 8 lock to 4.15GHz and stay there.

You can use Ryzen Master to adjust the multiplier when using PE 3 or 4. If you're booting at 41.5 with PE 3 then you can get it up to 42.5 and maybe higher.

https://www.overclock.net/forum/11-...tiplier-adjustments-through-ryzen-master.html

 

[Martin778]

Ryzen Master shows that it's hitting EDC limit of 140A under Cinebench R15 and resulting in ~4094MHz all core freq, when I increase it to 168A (max I can set in RM) it boosts to 4240MHz but results in 74*C Tdie under R15.

 

[nick name]

Ryzen Master shows that it's hitting EDC limit of 140A under Cinebench R15 and resulting in ~4094MHz all core freq, when I increase it to 168A (max I can set in RM) it boosts to 4240MHz but results in 74*C Tdie under R15.

What cooler do you use?

And have you set a VCORE offset in BIOS?

 

[Keith Myers]

hey keith! i used bios flashback from 0509 to 1201 and back to 0509. So yes it's possible. Hope that answer helps you.

Gesendet von meinem SM-G950F mit Tapatalk

Thanks I was worried because I think the WMI BIOS' make changes to the SMU firmware code and you normally can't revert back through those changes unless you use the AFUDOS app.

 

[Wuest3nFuchs]

Thanks I was worried because I think the WMI BIOS' make changes to the SMU firmware code and you normally can't revert back through those changes unless you use the AFUDOS app.

I cant tell you that exactly .
But i noticed the Firmware numbers changed in bios ,on windows 10 i have hwinfo running .
It showed me the asus ec and ite Controller before i've flashed the board the first time.
After the bios update it showed me the wmi style in hwinfo.

Gesendet von meinem SM-G950F mit Tapatalk