CCC 13.8 Causing Mouse Lag on Desktop?

[BradleyW]

Since installing these drivers, my mouse will lag and it becomes very choppy whenever I am loading a webpage or a program on the computer usually. The movement is very choppy and slightly lagging. Any suggestions?
This issue is only present when I have CCC 13.8 installed.

I use the Logitech G400 with Pointer Precision OFF, as usual.
Win 7 64-bit.

Thanks for the help.

 

[tunatime]

Ah this must be whats been hapning to me the last few days i thought my mouse was going bad. In games it works fine but on the desktop it lags and jumps

 

[Unknown-One]

From the 13.8 release notes:

Resolved intermittent mouse cursor corruption when rapidly moving the cursor across window borders

Last time AMD implemented a fix for cursor corruption, it caused issues with cursor responsiveness near the corners of all attached monitors. I have no doubt that this new cursor-corruption fix is having the same effect over a much larger portion of the display.

Previous drivers simply overwrite the cursor buffer with new data every tick, which is great for performance... but if an offset sneaks in, the cursor corrupts.

It looks like 13.8 is set up to either clear (or verify and clear if verification fails) the buffer every-other-tick when certain conditions are met. This is a brute-force fix that will prevent an offset from causing the buffer to corrupt, but it comes at the cost of cutting cursor performance in half.


Edit: Honestly, AMD is having so many issues here that they might be better off simply disabling hardware cursor acceleration. There's a flag they can implement in their drivers to tell Windows Vista/7/8 that they need to use software rendering on the cursor at all times (early intel drivers use this flag). As a temporary work-around you can open the Windows screen magnifier (and set magnification to 100% so it doesn't actually magnify the screen). The magnifier forces software cursor acceleration while it's running.

 

[BradleyW]

So this means AMD will always optimize the mouse in-game at the cost of desktop performance? If that is true, I will move to Nvidia.

 

[Unknown-One]

So this means AMD will always optimize the mouse in-game at the cost of desktop performance? If that is true, I will move to Nvidia.

it means that AMD is trying to keep image quality intact on the desktop, at the cost of performance (when conditions are met that trigger the new "fix").

 

[BradleyW]

What would you suggest I do? I want to keep the frame pacing but I do not want a poor performing mouse pointer.

 

[Unknown-One]

Like I said, start the screen magnifier (it comes with Windows). Even if it's not magnifying the screen, it forces software rendering on the mouse cursor.

That's pretty much the only work-around that exists until AMD fixes things...

 

[BradleyW]

Thank you very much for the support so far.
I will give that a try and report back.

Edit: Magnify works and removes the issue, however the mouse feels as though some form of Vsync is being used (Slight input delay, very slight). Anything I can do about that? I use a G400 and MS drivers.

 

[Final8ty]

Not getting any mouse lag.

 

[Unknown-One]

Not getting any mouse lag.

You're running 5970's. This particular cursor corruption fix is probably restricted to newer cards.

 

[Tanquen]

Older cards have the issue also and the fix did not fix it anyway. I got the mouse lag in spots now but the cursor corruption still randomly happens. I think it may be if you have two screens the same res you are good but anything else and you will get the corruption. The fix probable only works on certain resolutions and landscape/portrait combos. I notice mine happens a lot when I move up the bottom of the portrait display up onto the landscape display. My 5970 did the same kind of thing.

 

[Unknown-One]

Shoot down that theory, I've seen it happen on single screen and multiple (matched) screen setups.

Like I said, i think their best bet is to cut their losses and just forget about hardware acceleration on the cursor until they can figure out wtf is actually wrong. These band-aids are just going to frustrate people.

 

[BradleyW]

Well, I would like to know why AMD are so bothered with messing with the mouse, unless it has some sort of affect in game? Maybe reduced mouse lag?

 

[Tanquen]

Shoot down that theory, I've seen it happen on single screen and multiple (matched) screen setups.

Like I said, i think their best bet is to cut their losses and just forget about hardware acceleration on the cursor until they can figure out wtf is actually wrong. These band-aids are just going to frustrate people.

Sorry, just so used to being told that I’m the only one with the issue and that I must have a virus or put my computer together wrong.

Single screen? I’ve only seen it when moving the mouse from one screen to the other. It makes since there is some kind of transition that AMD has messed up when the pointer is switching from one screen to the other but on a single screen???

 

[Tanquen]

Well, I would like to know why AMD are so bothered with messing with the mouse, unless it has some sort of affect in game? Maybe reduced mouse lag?

It’s kind of a basic thing and should just work. Not everyone plays games and not all the time.

 

[Unknown-One]

Single screen? I’ve only seen it when moving the mouse from one screen to the other. It makes since there is some kind of transition that AMD has messed up when the pointer is switching from one screen to the other but on a single screen???

Yup, it can happen on single-screen as well, though it's less likely. Screen edges / corners can set it off, as can quickly running your mouse back and forth across window boarders (which causes the cursor to flip between regular and the grab cursor over-and-over).

It’s kind of a basic thing and should just work. Not everyone plays games and not all the time.

Agreed, AMD's issues with cursor corruption frustrate me to no end. Not even being able to use the 2D desktop without the possibility of the cursor suddenly going garbled is inexcusable after all these years and multiple change-logs calling it "fixed"

 

[BradleyW]

All AMD need to do is leave the mouse alone in 2D and optimize the hell out of the mouse in 3D! So, I'm using the windows magnify tool at 100% to fix the issue. However, there is slight input delay. Any way to reduce that?

Thank you.

 

[Unknown-One]

All AMD need to do is leave the mouse alone in 2D and optimize the hell out of the mouse in 3D! So, I'm using the windows magnify tool at 100% to fix the issue. However, there is slight input delay. Any way to reduce that?

What you're using right now is exactly what you'd get if AMD left the mouse alone in 2D. 100% software accelerated cursor handled entirely by the OS.

I don't get any additional delay with the magnify tool open, so I'm not sure how tel help you there.

The cursor-corruption problem is a long-standing issue on AMD cards, starting with the HD 5000 series. The HD 6000 and HD 7000 series share a similar architecture and have also carried the cursor corruption issues forward. I have a sneaking suspicion this might be a hardware problem that they're attempting to solve via software...

 

[djcoder]

Just installed the 13.8 beta on my Acer Aspire AM3450 EF10P, which is a prebuilt desktop with an AMD Radeon HD 7670 (it's hard to find reference to this anywhere on the internet, since it's an OEM-only card and people only talk about the 7670M) that I use for Hyper-V clients. I have no mouse lag on the desktop or in games - in either the master operating system or the virtualized operating systems (Windows 8, Windows Server 2012, and Ubuntu). I think that blows your new generation exclusively idea out of the water, but then again it might not - remember this card is OEM only so maybe they haven't applied the same tweaks.

I'm with BradleyW though, AMD drivers have always sucked in my opinion. This just reinforces my decision to go NVidia on my gaming workstation.

-Cameron

 

[Elmy]

Just installed the 13.8 beta on my Acer Aspire AM3450 EF10P, which is a prebuilt desktop with an AMD Radeon HD 7670 (it's hard to find reference to this anywhere on the internet, since it's an OEM-only card and people only talk about the 7670M) that I use for Hyper-V clients. I have no mouse lag on the desktop or in games - in either the master operating system or the virtualized operating systems (Windows 8, Windows Server 2012, and Ubuntu). I think that blows your new generation exclusively idea out of the water, but then again it might not - remember this card is OEM only so maybe they haven't applied the same tweaks.

I'm with BradleyW though, AMD drivers have always sucked in my opinion. This just reinforces my decision to go NVidia on my gaming workstation.

-Cameron

LMAO ... You think NVIDIA is better.... Go read their forums with all the problems with their drivers lately.

 

[stealthy123]

What you're using right now is exactly what you'd get if AMD left the mouse alone in 2D. 100% software accelerated cursor handled entirely by the OS.

I don't get any additional delay with the magnify tool open, so I'm not sure how tel help you there.

The cursor-corruption problem is a long-standing issue on AMD cards, starting with the HD 5000 series. The HD 6000 and HD 7000 series share a similar architecture and have also carried the cursor corruption issues forward. I have a sneaking suspicion this might be a hardware problem that they're attempting to solve via software...

I am running crossfired 7850s and I have zero cursor problems.

I have been reading a lot of your posts lately, and I would suggest you lay off acting like the problems you have with AMD are problems that are universally true. They are not, as evidenced by my not having any of the issues you seem convinced everyone has. It seems to me they may be specific manufacturing defects in individual cards, because if it was a design flaw it would be every single person having the issues you espouse. You really need to stop talking about problems with AMD cards as though every single one out there has issues, they don't.

People who don't know better are going to take what you say as an accepted truth.

 

[Tanquen]

I know but this is a known issue with AMD cards. My 5970 did it and my 7970s do it. It can be operator error and or a bad card for some issues but I have two totally different builds with different cards and both have the issue. I think it’s a hardware/driver issue that shows up under certain combinations of screens and resolutions. Like the random crashing of 7970s only on X79 motherboards. AMD support and others in forums just acted like you are crazy and that you don’t know what you are doing. Telling you to replace the RAM-CPU-MB-PS or you’re holding it wrong. When for almost a year AMD knew about the issue and told only few in a random forum I did not find until spending many hours swapping cards and motherboards. Then they release a fix for it and say that it was only in X79 Crossfire using eyefinity setups but it fixed my (at the time) single 7970 setup. How about the 7XXX DirectX9 issues? Same kind of thing. AMD support don’t know anything and all their cards fully support DirectX9 and plenty of people in forums saying that this and that DirectX9 game runs fine on their system. Then AMD released a fix for it that fixed some of the DirectX9 games.

I could see it being bad cards but then it’s a lot of them and AMD is trying to put driver band aids on hardware design flaws or large groups of poorly manufactured cards. Why aren’t they telling us to swap our bad cards instead of saying they fixed in a driver update?

 

[Final8ty]

You're running 5970's. This particular cursor corruption fix is probably restricted to newer cards.

I also have Trifire 7950's

 

[djcoder]

LMAO ... You think NVIDIA is better.... Go read their forums with all the problems with their drivers lately.

I've used NVidia ever since my first laptop. The only issue I've ever had with their drivers was Linux and AMD is only marginally better on Ubuntu.

My friend has 3 NVidia workstations - one with a GTX 690, one with a GTX 760 and the other one I don't know the card... GTX 2xx I think, or 4xx? Anyway, whenever I bring a computer around to his place for a LAN party or vice versa, he NEVER has driver issues. EVER. He always runs the latest beta available for the graphics card and he runs Windows 8 on everything. And I know he didn't have to mess around because I was there for 2/3 builds.

I'll tell you a story. Last time I was over at his house I installed Windows on his new GTX 760 build (his console replacement), downloaded the latest beta driver, and had no problems. The End.

Short story, right?

When you have an AMD card, oh no no no, you want your card to WORK? Impossible! Every update with AMD has one fault or another - including this one, GUIMiner is now screaming "Verification Failed, Check Hardware", so even though I'm immune to the mouse bug I'm STILL having issues and AMD's godawful drivers are actually costing my Bitcoins.

And the last one had a weird overscan bug where it detected my monitor as a cheapo television and turned on massive amounts of overscan, which did not disable no matter what in CCC.

I'm not trying to start ww3 here, every card has it's pros and cons, and honestly the biggest con of AMD is the drivers are just generally really bad and have issues no matter where you look. It's not the card, since I have two AM3+ mobos and two AMD cards (7670 and 6950), it's the awful drivers. And again, I'm NOT an AMD hater - they blow every NVidia card out of the water for mining - but their drivers are bad.

 

[Relayer]

I'm not running AMD right now, but I've had exactly one driver issue with AMD, and I used them for years. There were two driver builds in a row that when using "soft shadows" in advanced OpenGL settings with Cinema 4D the app would crash. I rolled back to the previous working drivers and waited for a fix, which as I said came 2 drivers later. This was with a professional modeling program using Radeon cards, so I couldn't really expect any support. I sent a bug report to AMD and they told me they didn't guarantee consumer cards on pro apps. but would look in to it. From my experience I have to conclude that people who have constant driver issues are doing something wrong.

 

[BradleyW]

I've been getting many people to test their mouse with the new drivers. Most of them all said that no issues were present. So, I decided to challenge that and I told them to do certain tests and look out for certain things. Now, most of them see the exact issue I described and they agree this does not happen on previous CCC versions. This shows people perceive things differently when trivial issues are present.
Why do we need hardware governed mouse control anyway?
Cheers.

 

[Tamlin_WSGF]

I've been getting many people to test their mouse with the new drivers. Most of them all said that no issues were present. So, I decided to challenge that and I told them to do certain tests and look out for certain things. Now, most of them see the exact issue I described and they agree this does not happen on previous CCC versions. This shows people perceive things differently when trivial issues are present.
Why do we need hardware governed mouse control anyway?
Cheers.

Why don't you post it here how to test for these issues?

Oh, and since logitech is known to be the source of some mouse lag in some games, could you try it with this fix as well? :
http://forums.steampowered.com/forums/showthread.php?t=1279444

 

[Tanquen]

Why don't you post it here how to test for these issues?...

Not sure why or if other programs do it but VMware Workstation 9 has major mouse lag over the Devices section middle left when you click one of the tabs for a particular VMware session. Was never there before.

It's not easy to see in the video but it's like an area of molasses and then the mouse will shoot off as you speed up a bit moving through the area. It's also still moving one direction after I've already started to move another. You can also see that the pointer is corrupted.

http://youtu.be/scmYRn9oCHY

 

[Unknown-One]

I have been reading a lot of your posts lately, and I would suggest you lay off acting like the problems you have with AMD are problems that are universally true.

I never said they were universally true, only that the cursor corruption bug has been around for three series now. People are STILL dealing with it, despite AMD calling it "fixed" multiple times.

And it's obviously a valid and recognized issue by AMD, they did just note it in the changelog for 13.8, after all.

You can also see that the pointer is corrupted.

http://youtu.be/scmYRn9oCHY

Yup, that looks familiar...

Seeing the same weird slowness issue VMWare Workstation 9 as well. Going to drop back to an older driver tonight...

 

[BradleyW]

Why don't you post it here how to test for these issues?

Oh, and since logitech is known to be the source of some mouse lag in some games, could you try it with this fix as well? :
http://forums.steampowered.com/forums/showthread.php?t=1279444

It seems the fix on steam did not work.
A good way to test the mouse pointer lag issues is to open a browser. As the page loads, run the mouse in circles within the open window.
Do the same for CPU-Z splash screen. Issue should show here.
Installing software and moving the pointer within that open install window also brings out the worst in this issue for me.
Cheers.

 

[naticus]

I was getting this 'desktop' mouse lag also with 13.8. It drove me bat shit crazy so I went back to 13.6b2.