Contribute
Register

Stork's MyHero II Build: ASUS ROG MAXIMUS Z370 HERO X - i7-8700K - AMD RX 580 - Thunderbolt 3

Well, it's not my OpenCore EFI, it's @pastrychef!
link
You have to fill some stuff and make a new USBport.kext.
Good luck!
This is an EFI created for this configuration (or rather for my configuration, which in any case is totally compatible with that of this thread, from which it originates and then has been updated)
In the EFI there are minimum interventions (I'm learning ... Radeonboost is missing for example) and it was created to be able to update Catalina 10.15.6 to Big Sur Beta and has also granted updates to even more advanced beta.
I don't use Big Sur as my main system, so I didn't dig into much, but on initial inspection it seemed to me that everything was working fine.
I would be grateful to those who use it if (in full spirit of this forum) in finding possible or necessary changes or improvements, they would also inform me and the others here, in order to make the configuration of this thread created so well by Stork, valid and performing for a long time to go.
It is still in verbose mode ...
For my needs (related to the video card) I used the FakeSMC and related to then add the kext related to the GPU that tells me the temperature of the VII, I say this for those who want to go back to virtualSMC.
There are to enter the data relating to the serial etc .. and the SMBIOS is 1.1.

Being on Catalina, as the main system, I am continuing to use Clover's EFI, but this one from OC loads faster, both Catalina and Big Sur, but until I decide to work on it and study it a little more I feel more comfortable with Clover.

Share problems/success and changes/improvements, thank you ...

greetings to all
Thanks for sharing those. Really interesting comparing how the EFIs are put together as they are quite different in terms of which kexts and drivers are included and also quite a few things in the config files are different.

I am still on Mojave so have stuck with Clover but on the latest version of that and have been using all the Acidanthera kexts and memory drivers for a long time now. I had planned to wait until OpenCore is out of beta before switching but will likely make the jump when I finally upgrade probably to Big Sur now. Most music software is finally compatible with Catalina, but I will have to get a new audio interface which is another reason for waiting as long as I can :)
 
Thanks for sharing those. Really interesting comparing how the EFIs are put together as they are quite different in terms of which kexts and drivers are included and also quite a few things in the config files are different.

I am still on Mojave so have stuck with Clover but on the latest version of that and have been using all the Acidanthera kexts and memory drivers for a long time now. I had planned to wait until OpenCore is out of beta before switching but will likely make the jump when I finally upgrade probably to Big Sur now. Most music software is finally compatible with Catalina, but I will have to get a new audio interface which is another reason for waiting as long as I can :)
It comes from the SampleCustom.plist which is located in the Docs folder of the Open Core base package and is modified to work on this MB, without many additions or analyzes to make it more widely compatible.
If you add that it still comes from knowledge being studied, you will surely understand that in contrast to the other it is much more minimal. It has the only advantage of being specific for this thread, and I can't tell you the price of what other disadvantages compared to the other one because many why and how I have yet to get there and probably never will.
But for this very reason I asked to share the changes and improvements here, to carry it forward a bit together, having the opportunity to learn and have a valid and updated reference for this configuration available for all those who want to switch to Big Sur or just switch to OC even on older versions of MacOS.

Come by when you try it and let us know ... bye.
 
I'm still on 1801. I forgot to look lately to update the BIOS. I'll update to 2402 within the next couple of days.
Have updated to BIOS 2402 from 2301 and everything is working fine still. In fact I don't know if it is the BIOS update or latest VirtualSMC SuperIO plugin which just came out, but HWMonitorSMC2 is now displaying far more accurate motherboard voltages and temperatures for my 9th get CPU.
 
Have updated to BIOS 2402 from 2301 and everything is working fine still. In fact I don't know if it is the BIOS update or latest VirtualSMC SuperIO plugin which just came out, but HWMonitorSMC2 is now displaying far more accurate motherboard voltages and temperatures for my 9th get CPU.
Updated the BIOS to 2402. No problems. :thumbup:
 
Have updated to BIOS 2402 from 2301 and everything is working fine still. In fact I don't know if it is the BIOS update or latest VirtualSMC SuperIO plugin which just came out, but HWMonitorSMC2 is now displaying far more accurate motherboard voltages and temperatures for my 9th get CPU.

Thanks for opening this theme, I was left with the 2302 necessary to be able to install the new processor (from 1801) and then I have not done any updates. However all the values worked perfectly for me, perhaps because I have always used FakeSMC and derivatives, instead of VirtualSMC.
Now I wanted to try the new VirtualSMC (also because it seems to be suggested for Open Core, and therefore it will be useful if things don't change) and so I also took the opportunity to update the BIOS to 2402 (again no problem, but I suggest a collection screenshot of the settings in the previous version).

Now I have a few questions:
the most important is related to the GPU sensor (Radeon VII): how can I detect the temperatures? GPUsensor.kext appears to only work with FakeSMC.

Then how do I manage the fans with the new HWMonitorSMC2?
I would like to create a curve for each one as I have done up to now.

Intel Power Gadget 3.6.0 panic and freezed everything (so much so that I had to force shutdown with the power button), then I installed 3.7.0 and everything now works, with both families of kext for sensors, but I still have to evaluate the actual need (at least here in Catalina).

A greeting and thanks again to you and to anyone who wants to give me suggestions, otherwise I will try to ask and look around again.
 
Hi @Stork and @c0c0p0ps
I'm here to open ask for a comparison with your experiences regarding updating the BIOS to version 2402.

I made this update at a particular time, when I'm also trying to figure out if I have a way to switch to VirtualSMC from FakeSMC (maybe I've already mentioned it), but for my needs I understand that it will be a problem.
However in this particular moment, of course, I am always with my eye (and ear for the fans) at temperatures, fans and sensors and their behavior in various conditions and I have noticed a big difference in temperature reading.
By big I mean that in idle there is an increase of at least 5/7 degrees detected and at full load the temperatures jump to values that are scary (and are certainly not true because I checked).
The term "jump" is also fundamental because they do it with a speed that would be impossible precisely for physics.
With the previous BIOS everything was normal and at full load the processor hardly exceeded 71 degrees and it did so by getting there with a speed that physics would accept and also speed aligned with the speed and temperatures that the video card also managed (which has a whole other way to detect temperatures), now there is a huge discrepancy in behavior and values that makes it clear that something in the BIOS, in the reading of the sensors is wrong.
Also because even by conveying the air that comes out of the radiator, nothing has changed compared to before and if my processor had worked in the past (and works enough) at the temperatures it says today it would not be very good, but certainly putting the hand on the radiator, I would have been scared much sooner and I would have run for cover, instead of replacing only the fan with two more powerful (and unfortunately noisy) which, however, have always kept the situation under control compared to the 8700K.

Obviously I verified that the change in readings had also occurred in Windows 10 because, however functional the kext used here on the Hacks, they are always "fixed" and not formalized and also guaranteed by the manufacturers, and that change in behavior also occurred in Windows 10, but I also add that even the temperature detected in the BIOS configuration has changed by those 5/7 degrees at rest, even without having loaded any operating system.

The same readings therefore result in Windows 10 and on MacOS with both FakeSMC and VirtualSMC.

Before going back to the previous BIOS for these obvious inconsistencies, I would ask you to check, even briefly, how your processors behave compared to previous BIOS and let me know.

I thank you and greet you.

a little but important edit:
A further proof of the fact that the readings with the current BIOS are not true is that even under benchmark (I used Cinebench 20) and therefore I load to the maximum even if the temperatures marked are by sure thermal throttling, during the tests there is no trace and the scores have also improved by a few percentage points, so the optimizations that Asus claims with this release seem to be there, but there is this reading problem.
 
Last edited:
Hi @Stork and @c0c0p0ps
I'm here to open ask for a comparison with your experiences regarding updating the BIOS to version 2402.

I made this update at a particular time, when I'm also trying to figure out if I have a way to switch to VirtualSMC from FakeSMC (maybe I've already mentioned it), but for my needs I understand that it will be a problem.

However in this particular moment, of course, I am always with my eye (and ear for the fans) at temperatures, fans and sensors and their behavior in various conditions and I have noticed a big difference in temperature reading. By big I mean that in idle there is an increase of at least 5/7 degrees detected and at full load the temperatures jump to values that are scary (and are certainly not true because I checked). The term "jump" is also fundamental because they do it with a speed that would be impossible precisely for physics.

With the previous BIOS everything was normal and at full load the processor hardly exceeded 71 degrees and it did so by getting there with a speed that physics would accept and also speed aligned with the speed and temperatures that the video card also managed (which has a whole other way to detect temperatures), now there is a huge discrepancy in behavior and values that makes it clear that something in the BIOS, in the reading of the sensors is wrong.

Also because even by conveying the air that comes out of the radiator, nothing has changed compared to before and if my processor had worked in the past (and works enough) at the temperatures it says today it would not be very good, but certainly putting the hand on the radiator, I would have been scared much sooner and I would have run for cover, instead of replacing only the fan with two more powerful (and unfortunately noisy) which, however, have always kept the situation under control compared to the 8700K.

Obviously, I verified that the change in readings had also occurred in Windows 10 because, however functional the kext used here on the Hacks, they are always "fixed" and not formalized and also guaranteed by the manufacturers, and that change in behavior also occurred in Windows 10, but I also add that even the temperature detected in the BIOS configuration has changed by those 5/7 degrees at rest, even without having loaded any operating system.

The same readings therefore result in Windows 10 and on MacOS with both FakeSMC and VirtualSMC.

Before going back to the previous BIOS for these obvious inconsistencies, I would ask you to check, even briefly, how your processors behave compared to previous BIOS and let me know.

I thank you and greet you.

a little but important edit:
A further proof of the fact that the readings with the current BIOS are not true is that even under benchmark (I used Cinebench 20) and therefore I load to the maximum even if the temperatures marked are by sure thermal throttling, during the tests there is no trace and the scores have also improved by a few percentage points, so the optimizations that Asus claims with this release seem to be there, but there is this reading problem.

I am not at my computer to do a detailed check right now but my findings were the opposite regarding temperatures and CPU ramping up.

I have not used FakeSMC for a long time now so only can compare my system from the change to recent BIOS versions. I also have no windows install so can only refer to the readings in BIOS and through HWMontiorSMC2 which works much better with VirtualSMC.

The previous 2 versions of the BIOS which introduced support for 9th gen CPUs have both always run my CPU with too much voltage for my liking - My 9900KS is 127w compared to 95w for the regular 9900k so is quite power hungry!

I spent a lot of time messing around with BIOS configuration to lower my temperatures when I installed it and have now actually settled on simply keeping everything set to auto! I don’t even enable XMP for memory as it pushes too much voltage through the VCC. Fortunately, my RAM automatically is picked up at the correct speed. My CPU now keeps its VCore voltage down to about 1.3 max under load when my CPU turbos to 5gHz and normally about 1v when idling. With XMP on and extreme tweaking etc I was up over 1.4v regularly with temperatures up at 75 degrees! Now temps under load are a more comfortable 67-70.

I honestly tried so many different tweaks and settings in the BIOS based on various forums I read until I found one saying to just reset the BIOS to factory defaults and start from there which I did and have not changed since.

My CPU is also only air cooled with a Noctua D15 (duel fan) and in a fractal define r5 case I can barely hear it spin up under load. I have a Pwm exhaust on the back which runs from the motherboard then 3 intake fans on the front and base running from the case fan controller which is on its lowest setting which seems to pull in enough air for the cpu fans and exhaust.

Going from BIOS 2301 to the latest I have seen no difference in temps or performance. In fact I see more accurate voltage readings now through HWMonitorSMC2 - it has an entire section now with all the readings from the BIOS such as VCore, VCC, DRAM etc.

Seeing as you also have a 9th Gen CPU I would maybe try a complete reset on your BIOS then only change Fast boot to disabled, enable Above 4G decoding, and enable your onboard GPU if you are in an iMac system definition - I run 19,1 for the 9900KS with iGPU enable as headless and still use Clover even though I run all the Acidanthera kexts and memory drivers.

As for the Radeon VII, I really cannot comment as I just have the Sapphire RX580. I am sure there are plenty of others using one of them on a Z370 board though.
 
I am not at my computer to do a detailed check right now but my findings were the opposite regarding temperatures and CPU ramping up.

I have not used FakeSMC for a long time now so only can compare my system from the change to recent BIOS versions. I also have no windows install so can only refer to the readings in BIOS and through HWMontiorSMC2 which works much better with VirtualSMC.

The previous 2 versions of the BIOS which introduced support for 9th gen CPUs have both always run my CPU with too much voltage for my liking - My 9900KS is 127w compared to 95w for the regular 9900k so is quite power hungry!

I spent a lot of time messing around with BIOS configuration to lower my temperatures when I installed it and have now actually settled on simply keeping everything set to auto! I don’t even enable XMP for memory as it pushes too much voltage through the VCC. Fortunately, my RAM automatically is picked up at the correct speed. My CPU now keeps its VCore voltage down to about 1.3 max under load when my CPU turbos to 5gHz and normally about 1v when idling. With XMP on and extreme tweaking etc I was up over 1.4v regularly with temperatures up at 75 degrees! Now temps under load are a more comfortable 67-70.

I honestly tried so many different tweaks and settings in the BIOS based on various forums I read until I found one saying to just reset the BIOS to factory defaults and start from there which I did and have not changed since.

My CPU is also only air cooled with a Noctua D15 (duel fan) and in a fractal define r5 case I can barely hear it spin up under load. I have a Pwm exhaust on the back which runs from the motherboard then 3 intake fans on the front and base running from the case fan controller which is on its lowest setting which seems to pull in enough air for the cpu fans and exhaust.

Going from BIOS 2301 to the latest I have seen no difference in temps or performance. In fact I see more accurate voltage readings now through HWMonitorSMC2 - it has an entire section now with all the readings from the BIOS such as VCore, VCC, DRAM etc.

Seeing as you also have a 9th Gen CPU I would maybe try a complete reset on your BIOS then only change Fast boot to disabled, enable Above 4G decoding, and enable your onboard GPU if you are in an iMac system definition - I run 19,1 for the 9900KS with iGPU enable as headless and still use Clover even though I run all the Acidanthera kexts and memory drivers.

As for the Radeon VII, I really cannot comment as I just have the Sapphire RX580. I am sure there are plenty of others using one of them on a Z370 board though.

Thanks for the detailed reply. It took me a while because I wanted to deepen.
Although I hardly changed anything on the BIOS, I had cloned the XMP settings from the previous version, so first I went to put them back in Auto, but nothing has changed in the behavior of the readings and temperatures above all.
Then I also tried to do as you suggested, mostly for scruple, because there was no more settings than the basic ones that you also mentioned, but despite the passage nothing has changed in behavior.
So I went to windows again and did all the correspondence checks and everything remained the same there too.
All I had to do was go back to the BIOS I had before and so I did, but surprise: the problem remained there with the version that used to work fine for me so I got worried.
I got an excellent quality thermal paste (not that the previous one was bad) and I replaced the thermal paste, even if it hadn't been in place for long and in fact it was practically perfect.
With this new thermal paste the values at rest are back to being as they were before, even with the latest BIOS, and I have gained quite a few degrees under maximum load with Cinebench R20, but this is thanks to this paste which in fact is renowned for performing well and earning several degrees.
In any case, my temperatures under full load are far from what I had before, in which at most I reached 71 degrees and I don't know why, since not even a return to the previous BIOS has changed anything.
The fact is that I still believe that the temperatures indicated are not the real ones because the radiator does not overheat in any case (and if the water that circulates goes close to the temperatures it says, it should), and the behavior, in terms of speed of temperature change is exactly what I had described earlier.
So now for sure everything is working colder by several degrees (and that's good), but I don't have the exact pulse of the situation.
Also because it has always been the video card to be the fastest one to warm up when it works and to go higher in degrees, but now it seems to have become the quiet one.

I wanted to report because I would not have alarmed anyone with these findings on the new BIOS and since the problem remained with me even returning to the old one, obviously there are two things, or something has changed with the update which then remained unchanged (but I see it highly unlikely) or something is happening to my Motherboard and I will see developments shortly, unfortunately.

I would ask you and I would be very grateful, when you return to your computer to try at least once to launch Cinebench R20 and let me know what it detects, especially in terms of temperatures, so at least I will have a reference on an app that stresses the processor in the same way , although the cooler and processor are different, but in theory they should have been designed to run at more or less similar temperatures under full load.
In the meantime and later I hope that others will also check and let me know, so at least I will have a clear idea that my motherboard is abandoning me ...:banghead:
 
Thanks for the detailed reply. It took me a while because I wanted to deepen.
Although I hardly changed anything on the BIOS, I had cloned the XMP settings from the previous version, so first I went to put them back in Auto, but nothing has changed in the behavior of the readings and temperatures above all.
Then I also tried to do as you suggested, mostly for scruple, because there was no more settings than the basic ones that you also mentioned, but despite the passage nothing has changed in behavior.
So I went to windows again and did all the correspondence checks and everything remained the same there too.
All I had to do was go back to the BIOS I had before and so I did, but surprise: the problem remained there with the version that used to work fine for me so I got worried.
I got an excellent quality thermal paste (not that the previous one was bad) and I replaced the thermal paste, even if it hadn't been in place for long and in fact it was practically perfect.
With this new thermal paste the values at rest are back to being as they were before, even with the latest BIOS, and I have gained quite a few degrees under maximum load with Cinebench R20, but this is thanks to this paste which in fact is renowned for performing well and earning several degrees.
In any case, my temperatures under full load are far from what I had before, in which at most I reached 71 degrees and I don't know why, since not even a return to the previous BIOS has changed anything.
The fact is that I still believe that the temperatures indicated are not the real ones because the radiator does not overheat in any case (and if the water that circulates goes close to the temperatures it says, it should), and the behavior, in terms of speed of temperature change is exactly what I had described earlier.
So now for sure everything is working colder by several degrees (and that's good), but I don't have the exact pulse of the situation.
Also because it has always been the video card to be the fastest one to warm up when it works and to go higher in degrees, but now it seems to have become the quiet one.

I wanted to report because I would not have alarmed anyone with these findings on the new BIOS and since the problem remained with me even returning to the old one, obviously there are two things, or something has changed with the update which then remained unchanged (but I see it highly unlikely) or something is happening to my Motherboard and I will see developments shortly, unfortunately.

I would ask you and I would be very grateful, when you return to your computer to try at least once to launch Cinebench R20 and let me know what it detects, especially in terms of temperatures, so at least I will have a reference on an app that stresses the processor in the same way , although the cooler and processor are different, but in theory they should have been designed to run at more or less similar temperatures under full load.
In the meantime and later I hope that others will also check and let me know, so at least I will have a clear idea that my motherboard is abandoning me ...:banghead:
Hi. I just downloaded and ran Cinebench R20 and have attached a quick screen grab I took during the render pass. With all cores running at 5gHz my temps spiked up to 77 degrees but quickly came down to average around 73 which is similar to what I get doing an offline bounce of a busy pro tools mix session. Considering how quiet my machine runs normally cinebencz really made my fans spin up :) In terms of score I am a long way behind the 48 core machine!
 

Attachments

  • Screenshot 2020-09-15 at 19.48.45.png
    Screenshot 2020-09-15 at 19.48.45.png
    2.5 MB · Views: 76
  • Screenshot 2020-09-15 at 19.49.32.png
    Screenshot 2020-09-15 at 19.49.32.png
    239.8 KB · Views: 81
Hi. I just downloaded and ran Cinebench R20 and have attached a quick screen grab I took during the render pass. With all cores running at 5gHz my temps spiked up to 77 degrees but quickly came down to average around 73 which is similar to what I get doing an offline bounce of a busy pro tools mix session. Considering how quiet my machine runs normally cinebencz really made my fans spin up :) In terms of score I am a long way behind the 48 core machine!

Thank you for taking the test right away and sharing the results.
Yes, Cinebench R20 is a remarkable test for the processor, it sterss it even more than what happens with a rendering created with a 3D program and I don't know how it does it, but I think it's because it is optimized to ask only that job and therefore the answer. it is even more targeted and the resources that the system reserves are the maximum possible. The fact is that the fans (if everything is set well) go to maximum, because the temperatures do the same and you realize how noisy the PC can become ...
Seasonal temperatures are probably also influencing the results compared to what normally happens and probably if you had carried out the test in a less hot period you would have obtained those 71 degrees that I did too.
However your air cooler for that 130W+ processor really does a great job! (it should investigate the stability over time, but if you do not need it, I advise you not to use it all unnecessarily, because in the best case scenario you come out with a nice cooking of the thermal paste).

I salute you and I promise you that you will be the first to be informed when my motherboard will give me some nasty surprise ;), but for now also given the heat, I avoid continuing to try to understand more, because so much going around in circles while probably aggravating the situation.
We'll see how long it lasts and in the meantime I hope others will tell me what they get.

In summary I understood that if I want to continue reading the temperature of the dGPU I cannot leave the FakeSMC family and that for the rest when I load the working processor, more than at temperatures, I have to look at the frequencies and see if thermal throttling appears, and I would be really curious to understand at what temperature indicated by my motherboard this happens ...

I salute you and thank you again for your kindness.:)
 
Back
Top