Hello, everybody!
I'm having issues with the newest version of the AMD Radeon Software, which is currently 18.5.1. I decided to reinstall the OS from scratch because of some issues with the computer itself and for some cleanning up, then I decided to install the most recent drivers of my devices (because I don't trust Windows 10's drivers update). The moment I install the AMD Radeon Software, my computer freezes for 1 second every 5 seconds I move the mouse, it's incredibly annoying and I can't figure out why. I restarted multiple times, but the issue is still there, so I decided to uninstall it. Once I restart, it works just fine, no stuttering nor freezing, even Windows 10 "updated drivers" work without any issues. But then again, whenever I play a game, it recommends me to update the GPU drivers.
To give you an idea, here's a gif'd video I recorded with my phone on a pitch-black picture and the cursor, drawing small circles and capturing every single stutter:
Despite happening with the cursor itself, it actually freezes my entire computer. I played a video and the moment I move the mouse, EVERYTHING gets stuck for a second and so on. I've checked my PC performance and there's no process consuming the entirety of the CPU, RAM nor GPU, and it happens right after I install AMD Radeon Software 18.5.1. I haven't tried with a previous version, but if there's anything I can change on the settings to fix this, please, let me know, this issue is very annoying.
My PC specs are as follows:
- OS: Windows 10 Pro x64;
- CPU: Intel i5 7600K 3.80GHz;
- GPU: ASUS Radeon Rx 580 8GB;
- Motherboard: Gigabyte H270-Gaming 3;
- RAM: Corsair Vengeance 16GB (2x8GB) 3200MHz;
- Storage: Corsair MP500 M.2 240GB + WD 3TB 7200RPM.
If I'm missing anything, please, let me know and I try to provide what you need as soon as I can.
Thank you very much in advance!
Yes, disconnect the cable from the port. Since even if you power it off, if it is connected to the port, the VGA still assumes you are using it and takes it into account. It's just you who do not see anything on the second monitor.
I have this issue as well since updating to 18.5.1. I was on 18.3.4 previously.
I have attached my DxDiag in a zip (pm'ed you the password).
So I have 3 monitors, as long as I have only one monitor enabled there is no issue. Works fine.
With multiple monitors I see the issue explained above. The whole graphics output freezes briefly every couple of seconds, obviously makes the system unusable.
When I "disconnect" my additional monitors this is done in the Display settings in Windows. I don't need to physically unplug anything.
Also it doesn't seem to matter which monitor I have enabled, if there is more than one it starts freezing.
The DxDiag should have all my system info. I will mention, I'm on Windows 10 LTSB, so I have security updates but none of the crap "features" Microsoft keeps adding to Windows 10 , like that Focus thing.
Also no GPU activity during the freezes that I can see with procmon.
I will be installing older drivers until this is resolved.
Thanks for all the reports folks, we were able to successfully reproduce the issue based on your feedback.
Hello,
I just like to clarify the responses made earlier in response to:
amdmatt wrote:
When folks say we turn a monitor off, do you mean press the power button or physically disconnect it from the back of the GPU?
First, before I start, I am awfully glad that this is not a hardware failure on my part and this issue has affected many other folks. Including the response above that the developer has been able to successfully reproduce it in a lab that this is a driver bug (or perhaps a Windows bug?).
Key points I believe the developers should know so that this issue can be fixed for everyone and not overlook the issue I am having. I have a dual monitor setup here and I experiencing mouse stutter accompanied by a micro system freeze when either display is powered off.
1. This issue affects users who are on multi display setup with their GPU's
2. This issue is present when all monitors are connected to the GPU and either display is turned off by pressing the power button on the display itself.
3. The issue is not present when only one display is connected to the GPU.
However I'd like to point out differences between users on this thread. One monitor is a DELL VE228 connected via HDMI and the second monitor is a ASUS SE2717H/HX connected via DVI to DisplayPort adapter (Unknown). The problem is persistent on the DELL while it disappears on the ASUS display after a small period of time on 18.6.1 driver with no workarounds implemented.
Thank you.
Ihorace.
Same thing here. See When extended screen is off mouse/graphics stutter.
I was just trying 18.6.1 this morning and issue is still present.
Back to 17.12.1
Hardwood wrote:
Ihorace.
Same thing here. See When extended screen is off mouse/graphics stutter.
I was just trying 18.6.1 this morning and issue is still present.
Back to 17.12.1
Yeah, the issue was introduced some where in the 18 series of drivers. For a minute there, I thought the problem was just with my hardware. For now, downgrade to 17.12.1, enable mouse trail (which doesn't completely erase the problem), or bare with it till it's fixed Which forces me to leave a display on needlessly or reboot to Linux.
im using 17.12.1 i dont need to have mouse trail on that version
Hey, just wanted to clarify, I recently updated to the latest driver, am having the exact same issues as everyone described (system/mouse lags every 3 or so seconds), but I do not have a multi-monitor setup. Just a laptop.
I ensured all my other drivers (including BIOS) are up to date and it made no difference.
I rolled back to an earlier driver and instantly the problem was gone.
Hello,
I've got the same problem as described, but with me it started with the 18.4.1 driver and now with 18.6.1 driver the problem still exists.
I also have a multi monitor setup with 4 monitors but until the 18.3.4 driver I didn't encounter any problems.
I haven't tried to only connect one monitor but i will try that later on.
But that cannot be the solution to only use one monitor because i need more than that.
I first thought it had to do with the new spring update from windows but when i put back my windows backup and i installed the driver 18.5.1 on the older version of windows the problem was the same.
I even tried a clean install with no better results.
Now I use the 18.3.4 driver and everything works fine but i hope you find a sollution before there comes a new game and i'm forced to update the driver to let it work.
I'm thinking about the new assassin's creed Odyssey, if the problem than still exists I maybe forced to buy another graphics card.
My system specs:
Asus ROG STRIX Z720H Gaming Motherboard
Intel Core i7-7700K CPU @ 4,20Ghz
MSI Radeon RX480 Graphics 8GB
32 GB DDR 4 2400 Mhz System memory
1 Samsung SSD 960 EVO 500GB
1 Samsung SSD 850 EVO 500GB
1 Seagate HD 3TB
2 Hitachi HD 2TB
1 LG Ultrawide 34UC79G Gaming Monitor with freesync (Connected to DP)
3 LG 24MP76 Monitors (2 connect to HDMI 1 conected to DP with cable DP-HDMI)
Logitech G502 Mouse
I have the same problem and it is very annoying! Since I updated the drivers to version 18.6.1 mouse lags and also, every x minutes, screen freezes and crash for one second! Please solve this bug! Btw can anybody tell me how to reverse the drivers to a previous working one?. Thanks.
Victor, Google DDU or Display Driver Uninstaller.
Install it and make sure it has a desktop icon.
Hold SHIFT and choose to RESTART.
When it restarts choose safemode no networking
When desktop appears run DDU and uninstall AMD.
Restart when done, install old drivers (that you've already downloaded).
I also have the same problem, I have a RX480 8 gb MSI, at the moment I am using the version of drive 18.3.4, after that all are giving freezing problems in versions 18.4.1 / 18.5.1 / 18.6.1. I hope the next update corrects this problem!
It might help if everyone reports this issue. Radeon settings/home/preferences/report issues online.
Hello, it's not going to make them move faster, they already acknowledged the problem. Perhaps they postponed until the next Driver release. Although it would be nice to a have a hotfix now or at least revert whatever code that introduce the problem until a fix could be applied later.
Well.. that's debatable. After acknowledgement of the Elite Horizons issue, it only took 3 or 4 driver versions to fix it.
Hey Guys,
youre problem is the closest I can Find to my one... thats Why I write in here.
My setup is:
Asrock B85M Pro4
Intel i5-4690
16 GB Ram
MSI RX480 8GB Gaming X
My Problem is that it seams to Buffer or something.
While gaming it runs normal for like a second, then it slows down to freez, than speeds up again. and slows down again. It keeps Fluctuating.
AMD MSI RX 480 8G Gaming x Fluctuation. Pt II - YouTube
Or
AMD MSI RX 480 8G Gaming x Fluctuation. PT I - YouTube
It runs fine with the old drivers like 16.2.... But I´d like to have tthat overlay option of relive and adrenaline....
pls help
One more strange thing - almost every time I finish playing the "Broforce" the mouse trails gets disabled and the bug appears again.
It would be really nice to see a proper fix soon.
I'm having this issue with 18.6.1 driver as well, before, I thought it was a problem with my HD, what I discarded because when using the onboard graphics the OS works fine without the freezing happening.
It may not be exactly the same problem as the other users because the freezing/flickering occurs even with only one monitor and even before login screen, but it's certainly something related to the compatibility of the current drivers and the windows 10 1803 build.
Additionally, I noticed that most of the time the freezing occurred (in my case) there was an usb connection / disconnect sound . So I used a PNP monitor to check what was going on and saw that it had a non-existent monitor device connecting and disconnecting. (see PnPMon64.txt)
Nirsoft's multiMonitor Tool also detected this non-existent monitor(see Monitors List.html). I used this to disable that monitor device and for the last 2 hours the freezing didn't occur.
PS.: For more information I'm including attached the dxdiag log and the windows event viewer log (filtered by errors).
edit: The freezing happened again. Updating PnPMon64.txt...
edit2: Happened again and now, intead of two, several monitors appeared in the MultiMonitor Tool.
I have this issue, however it only occurs when I plug in my 3rd monitor.
I have a Sapphire Nitro+ RX 580 4GB, and this issue only happens when I try to connect my 3rd display though an HDMI to DVI cable (have eliminated the cable by testing 2 different ones)
4690K
Sapphire Nitro+ RX 580 4GB
Gigabyte Z97N-Wifi rev 1.0
8GB 1600MHz Crucial Balistix Sport
Logitech G303 Mouse
I have the problem on WX 4100 pro card. when the second display is turned off, freezing is making my computer very annoying to use. but if I turn the second display on, there is no problem.
I found a solution that works for me..whenever I have the second display ( TV) turned off, I can press WIN + P, and choose windows setting "only one display" and the problem will go away. And if I want to use TV display, then I turn it on and activate second display also with the same WIN+P setting.
of course not ideal but this workaround might help someone else
I've been having this same issue, but only just read through this thread. Until today I'd simply been rolled back on a much earlier driver 17.x.x. I'm a software engineer and recently began working on a Vulkan based project, unfortunately the Vulkan support in the earlier driver is out-dated / broken, and so I had to update.
My system is an i7-6700k skylake running Win10-64 and an MSI branded RX-580 with four monitors attached (3xDell 2007FP and 1xLG FullHD on the HDMI). I also have an Oculus rift attached to the same RX-580.
This issue is more than just a mouse pointer problem, it appears to stall all video activity for around half a second every 2-5 seconds. Therefore, streaming video freezes (netflix / youtube), as do games, and of course the mouse pointer. It essentially makes the machine unusable.
As I had no other option but to roll back, and I couldn't roll back to major version 17 due to Vulkan, I decided to walk back through the drivers to identify which was the last that worked correctly. I downloaded every file available here: https://support.amd.com/en-us/download/desktop/previous?os=Windows%2010%20-%2064 and (along with the other versions that I already had up-to 18.6.x) began clean installing back.
Luckily, I didn't have to go back too far, 18.3.3-March18 worked for me. For anyone trying to find the latest driver that doesn't have this issue, I think it's worth starting there.
As a software engineer myself, I know that it's all too possible for these types of problems to be resolved, only to re-emerge in newer builds. Therefore, it's possible that this problem existed before 18.3.3 and re-emerged after it, so if you're stuck on a 17.x.x version, it's still potentially worth trying 18.3.3.
Wow, thanks for a bit in-depth look on things. It's always interesting to hear the opinion of someone who is working in this exact field.
But yes,as I wrote in my post when benchmarking this "hiccup" caused the GPU load to go from 90-100% to 0 for a second or so.As if the drivers crashed and started up once more. Since I'm using the card for work and mining,I noticed that when mining, in MSI afterburner, instead of having somewhat of a constant power draw,it fluctuates.Like it will be 102W and it will pop up to 120-130 every 2-3 seconds.It's interesting that while the power spikes,the clock frequency drops(I suspect the GPU load to 0% is caused by this). But I'm having issues with my computer ever since I installed the MSI RX570,so that might be related to the motherboard. I'm expecting a new motherboard and hoping that my problems with the constant freezes will be fixed after that.(but I suspect this is not due to the VGA) Since it freezes with my wold VGA as well.
One thing is interesting though since you mentioned you are a software engineer that works with Vulkan. On the benchmark I ran,when I selected the Vulkan option,it did not have the drop to 0% GPU load.When using the OpenCL version I had this behaviour.
I also reverted to previous drivers and am now on 18.3.2 and both Vulkan and OpenCL work fine.
thanks for doing the legwork (clean installing previous versions) Craig.
Summary
It started with 18.4.1 driver and persisted with 18.5.1 and 18.6.1. Reverting back to 18.3.4 fixes the problem always.
Cause
So a software change in the driver from 18.3.4 -> 18.4.1 introduced the bug. Seeing the results here the problem lies probably in the software component for detecting or handling multi-monitor setups. Maybe some kind of new sleep modes.
Fix
The driver department should take a look at their diff's for the software revisions and see the changes that were made.
DDU v17.0.9.0 in safe mode -> v18.7.1 !
This fixed the issue for you?
Try it and tell us, i dont have this issue at all.
I'm afraid to say that it still doesn't work for me.
At least they added this issue in the "Known Issues" section of their release notes.
juliannb wrote:
At least they added this issue in the "Known Issues" section of their release notes.
I'm not seeing it in the known issues.
https://www.amd.com/en/support/kb/release-notes/rn-rad-win-18-7-1
Known Issues, 4th note: "Cursor or system lag may be observed on some system configurations when two or more displays are connected and one display is powered off."
I don't know what they're talking about with the "one display is powered off", but ok.
They mean if you turn off the monitor the lag starts to happen.
Turning on Mouse Trail in Windows seems to slow the lag down
Well, that's my case.
My PC is connected to 2 screen, 1 is my TV, and one is my screen monitor.
If i am using the PC and the TV is turned off or in another source then i have the issue.
Hi all,
I now installed the new 18.7.1 driver and it solves nothing.
This is the first time I installed a new driver since my last message and i now also found out that it has to do with a monitor turned of.
I have 2 pc's and 4 monitors.
My main pc with the amd Radeon RX480 card is connected to all 4 monitors and my old pc is connected to one of them.
In windows and in radeon settings they are numbered monitor 1, monitor 2, monitor 3 and monitor 4
When all monitors are turned on and they all show the screen from my main pc with the amd card, there is no issue and everything works fine.
But when I turn off the power from monitor 2 or monitor 3 or when I switch monitor 3 to show the screen from my second pc then the freezing starts to appear every few seconds.
It's a screen freeze because the sound goes on but the video freezes for a second and that for every few seconds. (this happens on all monitors that are turned on)
But when I turn off monitor 1 or monitor 4 or 1 and 4 both together I don't have the freezing issue.
Update:
I now pulled the plug form monitor 4 and what happened is
All 3 monitors on no problem, 1 or 3 turned off no problem, but 1 and 3 on and 2 turned off it freezes again.
I also found out that when I turned off 3 and 4 together and leaving 1 and 2 on there is no problem.
And when I only leave on monitor 1 there is also no problem.
And last when I turn off 1 and 2 and I Leave 3 and 4 turned on the freezing starts again.
Maybe there are more options but i think this is more than enough input for one day.
So as i see it in my case, it seems that when the chain of monitor's is broken in the middle the freezing starts and other wise it works good.
For now the new driver can stay because i have figured out how to work with it, but it's not ideal.
So please try to fix it so we can turn on and off any monitor we like.
Is there anyone here who has the same experience as I have?
I have two monitors... both on all the time.. lag exists. Primary is a new-ish AOC 2879 28" 4k. Secondary is an OLD 22" AOC V22 @ 1680x1050.
Turning on mouse trails helps.
Going back to older drivers *18.3.x" helps..
juliannb wrote:
https://www.amd.com/en/support/kb/release-notes/rn-rad-win-18-7-1
Known Issues, 4th note: "Cursor or system lag may be observed on some system configurations when two or more displays are connected and one display is powered off."
I don't know what they're talking about with the "one display is powered off", but ok.
Oh I see, I was being shown the 18.5.1 notes.. seems a little backwards the way they have it. Thanks for pointing it out.
Yes, the one display (of 3) being powered off is when mine exhibits this.
Problem extisted for me a few driver releases back and was gone until today. I installed 18.7.1 :-(
Monitor is connected via dvi and on HDMI is the Sound-Receiver connected ...
Hello, i just updated to 18.7.1 drivers from 17.x drivers and this exact issue appeared...
I've downloaded the process explorer and noticed that everytime the spike occurs the "System" process jumps to ~3.5% CPU usage, no GPU usage spikes seen...
I have 2 monitors (types are in the speccy screenshot), one connected to DVI-D and other one using DP to VGA active adapter
Mouse is just a generic chinese "gaming style" mouse..
I downgraded to 18.5.1 and everything is fine again.
Check out the irony: "PRO Hi-Tech" Russian Youtube hardware news channel just announced that they've got email spam from AMD about how some "QA Consultants" have "tested" AMD driver for being "most stable graphics drivers in the industry"… I've never understood why the hell would you need to advertise driver, of all things, necessary and irreplaceable part which you HAVE to provide for your hardware not to be a useless pile of junk. But doing so, while notoriously being the most incompetent and irresponsible in software engineering, is just another level in being delusional.
HIRE MORE PROGRAMMERS and LESS TALKING SUITS, you lying and censoring scumbags !
And, yes, you got it right: I also have RX 580 with 2 monitors attached (DVI+HDMI-to-DVI). Good thing, I use GNU/Linux as primary OS and AMD F/OSS developers are not as useless.
EDIT: here, fixed it.
It wouldn't be the first time I see this, I've spotted this at a post on Reddit and I was forced to give them the link of this thread.
I can't tell for sure how much of this is true, but as I said earlier, at least they acknowledge this issue in the Known Issues in the 18.7.1 version update, which is more than an accomplishment for me.