[SOLVED] PC freezes after installing AIO cooler

Sep 19, 2019
9
0
10
I recently built a PC last March 2019 with the following specs:

Monitor: Viewsonic XG2701
Motherboard: Asus ROG Strix B450 F Gaming
CPU:AMD Ryzen 5 2600X
RAM:2x 8gb ddr4 3000 Patriot Viper Steel, CL15 1.2v
M2 SSD: 500gb Samsung 970
GPU: 8gb Sapphire RX590 Nitro+ OC 256bit
PSU:Corsair 750watts PSU, HX, HX750
Casing: CM Cooler Master MasterCase H500 TG, iron gray, atx, 2x200mm 1x120mm
OS: Microsoft Windows 10, Home Edition 64bit
Keyboard and Mouse: Razer Cynosa Bundle
Headset: Steel Series Arctis 7

Other components coming from old builds are:
250GB SSD
500GB HDD

I recently installed a Cooler Master MasterLiquid Lite 240, idle temperatures improved from 40C-50C to 30C-35C.

After the installation of the AIO Cooler, when I play games, in this case They are Billions, the screen freezes, blacks out and recovers. The timing and frequency of the freezing differs sometimes it occurs after around 15 to 20 minutes of continuous game play then occurs every 3 to 5 minutes to sometimes every minute, number of screen freezes varies from 3 to 5 in one hour to as high as 10 to 12 in one hour. System log sometimes show errors on “amdkmdap” and “display driver encountered an error and recovered” that seem to be around the time of the screen freeze.

There were no Windows and GPU drivers updates done prior to the installation of the AIO cooler, but to be sure that software issues are ruled out, I did a clean install of GPU drivers and checked Windows updates for any broken update installations.

I encountered the said freezing before with Rise of the Tomb raider and Shadow of War using the stock CPU cooler but not as frequent as this one when the AIO cooler was installed and it seems to be addressed when fixing the GPU frequency and setting the power factor to 50%. Steam overlay was also disabled.

I also installed HWMonitor Pro and logged the time the screen freeze occurs, based on the graphs, CPU and GPU utilization, power and voltage seem to drop around the same time of the screen freeze.

I am not sure if it is a power supply issue but 750W seems to be enough for me.

I will try installing again the stock cooler and see if the issue is addressed.

Anyone knows why this happens?
 
Sep 19, 2019
9
0
10
Remove all of your overclocks?
Remove the current RX 590 drivers with Display Driver Uninstaller and install the latest from AMD.
Already did that last night but problem is still there.

For the overclocks, it is only the following;

for the CPU was at 5% set from the BIOS
for the GPU it is OC by default, only adjustment that i made is the power limit is increased to 50%, this is based on the error i encountered during Rise of the Tomb Raider (this seems to solve the previous issue) but after the error now this setting resets. Also tried running with default settings in Wattman, PC still freezes.
 
Last edited:
Sep 19, 2019
9
0
10
Power plan is Windows balanced/high performance, or Ryzen Balanced?
DOCP enabled/disabled?

Do you have a spare gpu and psu to test with?
Power plan is Balanced, even with High performance, error still occured
I am not familiar with DOCP setting, i do not remember adjusting anythin related to it.

Installed my old HX 620 PSU, i still encountered the error

Based on HWmonitor, at the time of the error, paramaters such as Power, Temperature, Utilization for both the CPU and GPU all drop
 
Last edited:
Sep 19, 2019
9
0
10
Just an update. Installed my stock cooler again, error occured still having the error below;

"Display driver amdkmdap stopped responding and has successfully recovered."

So I decided that this might not be a hardware issue, so did a fresh install of windows, while on going with installation of other software, encountered same warning as above
 
Last edited:

Phaaze88

Titan
Ambassador
DOCP is the XMP feature on their Ryzen motherboards - I've no idea why they had to be different here...

Windows was reinstalled... < That tends to fix a lot of software issues - before adding your 3rd party software, anyway... but, didn't work.
Swapped psus... didn't work.
Power plan... also didn't work

The reason those parameters drop during this error is because it causes the system to pause.

There are still 2 other hardware possibilities that come to mind:
The cable from the monitor to the graphics card - do you have a spare cable around to test with?
The graphics card itself... are there any on-screen artifacts, lines, or odd color changes when these 'hiccups' occur?
 
Sep 19, 2019
9
0
10
DOCP is the XMP feature on their Ryzen motherboards - I've no idea why they had to be different here...

Windows was reinstalled... < That tends to fix a lot of software issues - before adding your 3rd party software, anyway... but, didn't work.
Swapped psus... didn't work.
Power plan... also didn't work

The reason those parameters drop during this error is because it causes the system to pause.

There are still 2 other hardware possibilities that come to mind:
The cable from the monitor to the graphics card - do you have a spare cable around to test with?
The graphics card itself... are there any on-screen artifacts, lines, or odd color changes when these 'hiccups' occur?
I did not install any motherboard utilities just the drivers, clean install of latest Radeon software (19.5.2) but still I encountered the errors at two instances;
  1. while installing AutoCAD
  2. while using powerpoint (inserting a picture)

I will try if replacing the displayport cable will do anything

As for the visual indications during the 'hiccups'' the screen just freezes (mouse pointer dose not move) then screen blacks out for a few seconds, then returns to normal sometimes i see a no signal indication on the monitor if the black out duration is long
 
Sep 19, 2019
9
0
10
If changing the display cable doesn't work... consider an RMA, but even if you get that card replaced, the replacement 590 might do the same crap.
https://www.newegg.com/sapphire-radeon-rx-590-100415nt-8gsel/p/N82E16814202327 (check out the 3, 2, and 1 star reviews, some of them mention the same crap)

Also, what are the gpu temps like in a game?
Highest recorded temps is at 60C, it pretty much occured around the time of the error somehow if fluctuated from 45C maybe during one of the 'hiccups';

I also observed that some of the 'hiccups''occured with the GPU fan not even running, meaning load and temps are still low. I think the software, has some issues

I am thinking if possible of just installing the drivers and not the software (profile wattman), do you think it would work?
 

Phaaze88

Titan
Ambassador
No. How is wattman interfering with the graphics drivers?

The gpu core is running cool, at least. But the same may not be true of the memory chips or the VRMs.
The only way to figure those out without the proper equipment is just to monitor the 2 with Msi Afterburner's real time graphs:
-memory clock speed - which almost never drops unless it's overheating.

-for VRM, monitor gpu clock speed, Power level(%), and Power Limit
The card will clock down a little bit(20mhz or something) if it hits the power limit. If the VRMs are overheating, the power level and gpu clock speed should take a massive dive.
 
No. How is wattman interfering with the graphics drivers?

The gpu core is running cool, at least. But the same may not be true of the memory chips or the VRMs.
The only way to figure those out without the proper equipment is just to monitor the 2 with Msi Afterburner's real time graphs:
-memory clock speed - which almost never drops unless it's overheating.

-for VRM, monitor gpu clock speed, Power level(%), and Power Limit
The card will clock down a little bit(20mhz or something) if it hits the power limit. If the VRMs are overheating, the power level and gpu clock speed should take a massive dive.
I dont think it is worth the hussle, the card clearly has some issues, best route to follow is RMA.
 
Sep 19, 2019
9
0
10
No. How is wattman interfering with the graphics drivers?

The gpu core is running cool, at least. But the same may not be true of the memory chips or the VRMs.
The only way to figure those out without the proper equipment is just to monitor the 2 with Msi Afterburner's real time graphs:
-memory clock speed - which almost never drops unless it's overheating.

-for VRM, monitor gpu clock speed, Power level(%), and Power Limit
The card will clock down a little bit(20mhz or something) if it hits the power limit. If the VRMs are overheating, the power level and gpu clock speed should take a massive dive.
I do get the message "Default Radeon WattMan settings has been restored due to unexpected system failure" at time of the hiccups. so I was thinking that Radeon software is being unstable and causing the errors.

I recently installed Sniper Elite 4 to test the issue, radeon drivers is at version 19.4.3, when the game starts screen goes black and goes back to desktop with error message "d3d11 lost device error hresult 0x887a0005 reason 0x887a0006"

I did a clean install of the latest radeon software 19.5.2, I was able to open and play the game at ultra settings for a few minutes, I will try again in the next few days to see and maybe monitor using MSI afterburner

It is just sad if I will go the RMA route as all the parts are brand new when I built it last March 2019