Equipped with 18.8.2, for several days now, I have noticed in Speccy and Hwmonitor that my graphic card loads at full in memory and clock (2000Mhz and 300 Mhz) while there's no activity (0% according to Wattman profile from AMD Radeon software settings).
So even if I do nothing on my PC and let it runs alone, my video card loads at full and my temperature stuck between 50 and 53 °c all the time.
Yesterday evening, I have uninstalled the driver of my video card with DDU and reinstall the latest version of driver (18.8.2) but the problem remains after that, so nothing has changed at all.
I suspected Intel optane memory (which is installed and activated on my system), but after I have disabled it, the problme is still the same. (Intel® Optane™ Memory—Revolutionary Memory )
Then I have renamed amddvr.exe and amdow.exe and end theme from task manager, again, nothing has changed.
I have lastly uninstalled speccy, then Hwmonitor. Still remains after that.
So after I did all these steps, what can I do next?
My PC (Desktop System) :
I suspect the graphics controllers installed with your card is to blame. All third party controllers conflict with AMD settings and controls.
Delete Asus Tweak and Fan controls. Disabling them does no good. Delete then clean install the AMD drivers. Clean Install AMD Graphics Drivers
Thank you for your answer.
I forgot to mention it, I have not Asus Tweak GPU and not FanSpeed on my PC.
And as I said before, I have uninstalled and reinstall correctly the latest driver of my video card by following this procedure: Clean Install AMD Graphics Drivers
Apart from Speccy and Hwmonitor uninstalled, I have no similar software.
Also I disabled onboard graphic (Intel HD graphics) on BIOS a few weeks ago.
What can I do next?
Trying to eliminate possible causes...Re-enable the Intel graphics. Go to Intel support, run their "Driver Search" program and check for updates. See if a update changes anything.
Have you changed anything in Wattman? Power Limit and 'Max' temperature are the only things to change...the rest left at default (automatic)
"Trying to eliminate possible causes...Re-enable the Intel graphics. Go to Intel support, run their "Driver Search" program and check for updates. See if a update changes anything."
Actually, the default setting for onboard graphic is set to "AUTO" so since I use AMD GPU as primary video card, Intel HD graphics is automatically disabled. Whatever I set to "AUTO" or to "Disabled" would change anything.
And no, enabling onboard graphic has not changed at all in my situation.
For Intel driver & software update, there's no new version detected so my PC is updated for drivers and softwares (including Intel Optane memory).
"Have you changed anything in Wattman? Power Limit and 'Max' temperature are the only things to change...the rest left at default (automatic)
That's exactly I did on AMD Radeon settings, I set the limit power to max (+50%) and let everything else at default (including temperature).
I would roll back to the previous drivers Radeon™ RX 580 Previous Drivers | AMD
I would double check Windows/Services for any mention of Asus services.
Report the issue to AMD > Email Form
Okay, I think I finally figured out the problem.
It has been detected by gamers/users before the update 18.8.1 and AMD has taken it as a known issues => Radeon™ Software Adrenalin Edition 18.8.1 Release Notes | AMD
But it's not fixed in the current driver 18.8.2 yet (we don't know why it's not mentionned as a known issue => Radeon™ Software Adrenalin Edition 18.8.2 Release Notes | AMD
In order to be sure of that however, I want some evidences from the AMD community if this problme has been resolved in the 18.8.2 version.
In any case, I don't be worry anymore on my PC because I was trying to uninstall any suspicious software (speccy, CPUID HwMonitor etc...) while this is not the cause.
So time to update this topic.
After I updated my Windows 10 (October 2018 update with first culumative update) and installed the latest driver of my graphic card => 18.10.1 (Radeon™ Software Adrenalin Edition 18.10.1 Release Notes | AMD ), I see that it is not resolved (like AMD RX vega series).
So the problem is not the Radeon Wattman memory and frequency values, it is always high even in idle mode anyway.
So apart from RX vega series mentionned for elevated memory clocks (and memory I suppose), RX 580 (at least) is absent.
Unless I still have an isolated issue (I don't want to believe that).
Does someone has the same problem or not ?
Just the usual suspects, disable hardware acceleration in browser, and/or
Also for troubleshooting, disable/turn off live tiles/apps/websites that start minimised, and in task manager > start-up, disable all apps(not defender).
"Just the usual suspects, disable hardware acceleration in browser"
=> Okay, So I have disabled GPU rendering on Internet Options from Internet Explorer (although I disabled this web broswer from windows features) and from Firefox but nothing has changed, memory and frequency are still elevated at idle.
"Also for troubleshooting, disable/turn off live tiles/apps/websites that start minimised, and in task manager > start-up, disable all apps(not defender)"
=> I already did this, and the issue is still there.
Actually, I'm starting to think that the problem is optionnal drivers, maybe I'll have to wait the recommended driver later.
Does someone has the same problem or not ?
A lot of us have the 580 but can't use newer drivers because of the 5 second pause bug when using multiple monitors, otherwise we could check our systems too.
For you just roll back to working driver for now, but do send in the error report.
Although other members here don't agree, I NEVER use wattman etc. I make sure Radeon Settings are set to default (turns off wattman) and use MSI Afterburner. I never have problems.
"Although other members here don't agree, I NEVER use wattman etc. I make sure Radeon Settings are set to default (turns off wattman) and use MSI Afterburner. I never have problems.."
=> The only option I use is "Power consumption" to + 50%, and I let everything else to "default", I don't use the equivalent of "MSI Afterburner" (I think is "Asus ROG Strix" or whatever).
"For you just roll back to working driver for now, but do send in the error report."
=> I'd rather to wait a recommended driver later, it'll come out in October or November with the AMD RX 600 series, period.
Ever find a solution to this problem? I've been having it on a MSI Rx 580 8G model and it's driving me up a wall!
I had the same problem. I bet you have multiple monitors that run at different clocks. 59-60-75 for example.
If the speed of the monitors does is different, the gpu will always run at 2000mhz.
There is also a bug related to that, if all your screens are set to 60hz, but it still runs at 2000mhz do the following:
1)Rightclick on a free space on the desktop.
2) Display settings.
3)Scroll down to advanced display settings.
4) Display advanced settings for Display 1/2/3 (usualy its the freesync monitor or the one with the highest possible hz).
5)Display adapter properties for Display 1/2/3
6) Monitor folder. Make sure all the hz for all the monitors are the same.
7) Apply. Repeat for screen 1/2/3/etc till the mzh drops to 400 as it should.
On PC restart that error may return, the just repeat the process. Usualy its just 1 screen that produces that bug, make sure
to find out which by monitoring the mhz when doing steps 6-7.
Hope that helps
sometimes the display settings will show all the screens at the same HZ, but just apply a different one and revert back, on all your monitors. That will solve the 2000 mhz bug.
Welcome back guys, I am through to announce that my problem on my RX 580 has finally been resolved last week-end.
I have decided to switch my Acer KG271A to Acer XV272P and after installed it and connected to my PC in Displayport, the temperature displayed on the "Performance" tab on AMD Radeon Software control panel was finally real.
On idle, my RX 580 runs between 36°c and 38°c and when I play some Star Wars Battlefront II or Darksiders III, my graphics card goes up to 49°c or roughly 50°c and fater that => 38/39°c, pretty nice and impressive. No performance impact.
So, the problem was definitely coming from hardware : THE MONITOR.
I can say that it's officially fixed.
Thank you everyone for your contribution, I know it was too long and our discussion about my problem was very awful, however, in the end, it was worth to talk about it because I have finally found out why by myself.
Have a nice day and good gaming !