ASUS software is garbage (Update, 19-12-2022)

Hardware Hub GPU
1

I’ve been having some system problems lately and I even got a new SSD with fresh install of windows to try and solve it.

I get these system lockups where it’s not frozen as such… can still click on things and CMD will even open, but nothing actually happens. Can’t restart or shutdown either and after I power cycle the system my browser tabs are reset back like 30-60min. It reminds me a lot of how windows kinda stops working if the system drive loses power. You can click on things, but nothing really happens.

Anyway, I was suspicious of the ASUS GPU Tweak software as it seemed like I started having these issues after I installed it, but I saw errors logged in Event Viewer regarding NFTS errors which led me to believe that my 5 year old SSD might be on the way out even though SMART said all was fine.

I installed the ASUS garbage today for the first time since I got my new SSD just over a week ago and not 2 hours later, same lockup as before. No NTFS issues reported yet though.

OK, but why do I want/need to use this software?

My GTX 1660ti has 2 extra fan headers. I actually really like these as I can have 2 case fans completely stopped and not sucking in dust for no reason while the GPU is at idle or low load. But so far the only way I can control and set custom fan curves for these extra fan headers is with GPU tweak.

My motherboard has 4 chassis fan headers including the AIO header so I can with the use of splitters get all my fans setup and working the way I want using FanControl, but I would really to be able to use these headers if possible.

Anyone know of any existing script/app/anything that I can use to get control of these headers without using the garbage software from ASUS?

2

Update

Today I reconfigured my fans to all run on the motherboard using splitters on 2 of my fan headers. I was using FanControl to configure some curves and HWMonitor to verify things. And then I got a random freeze of the system EXACTLY like before. Then again while just watching a youtube vid while HWMonitor was open.

Since then I’ve made sure not to have any software that will check sensor info of the GPU open and so far it’s been good.

I believe that using any software to access sensor info for the GPU is triggering the issue. If I were to guess I would say it’s a driver issue. I updated just after the RTX4090 release as people were commenting on the DX12 performance uplift in general. I’ve updated a couple of times since, but the problem remains.

Next test is to use DDU and install a pre-4090 release driver and then see if the problem persist.

Side note, GPU Tweak is still garbage and I won’t go back regardless…

3

Something else that may be related is the fact that you can now check GPU usage on a per app basis in Task Manager.

How long has this been a thing? I’m not finding any info regarding when this was added, but I know it’s not always been a thing.

image

4

I believe it was added in redstone 3, it’s part of the new wddmv2 framework feature set expansion

1 Like
5

Thanks. I just installed the driver I had before the 4090 release which was/is 512.15.

Started up FanControl and HWMonitor and just letting them idle in the background. Now we wait and see…

6

I would think a WindowsKey+Shift+Ctrl+B when the freeze occurs could rule out or pin down the gpu driver being the issue.
Although to me is sounds like explorer.exe is crapping out and hanging.

7

The freezes are weird. The system doesn’t always completely freeze. While watching a youtube vid, for example, the rendering would just stop while the audio continues. I can’t exit/close any apps even though I can still move the mouse and the animations still work when clicking buttons. I can actually open the start menu and select restart or shutdown, but nothing happens.

I should point out that I had nvidia driver issues on my work desktop as well a few weeks back. That system is auto updating everything and it was auto updating the nvidia drivers, although not to the latest drivers. I then manually installed the latest drivers and that solves the bluescreens issues I was having at work.

But, the crashes at the office were actually logged in Event Viewer which is why I knew the issue was driver related. Here on my machine there is NOTHING in the event viewer regarding the freeze apart from the ‘unexpected shutdown’ caused by me having to reset the machine.

Another odd part of this freeze, is that it’s like the system goes back in time a few minutes. I would watch a youtube vid, close the tab when done, open a new vid, get the crash halfway through the new vid, and then the old tab with old video would be open after the restart. Sometimes it’s like 2-3 videos ago and it’s like what I’ve done in that time frame never happened.

Two decades of working in IT and the only time I’ve seen anything remotely similar was when we unplugged the power to a hard drive while windows was running. You can still click around and some things still kinda worked (like skipping songs in Winamp), but nothing you did during that time would persist… obviously.

I can’t exactly recall, but I think I installed GPU Tweak III soon after the driver update and so that’s why it looked like GPU tweak was the cause, but it’s looking like a driver issue.

I’ve seen a few other people online complain about freezing on their 16 series cards. Not enough info to make any sort of assumptions here, but I would have thought that if this was widespread we would all already know about it.

So far it’s been 30min with HWMonitor and FanControl running and it’s still OK. Earlier, on new driver, it went wrong in less than 30min. But I’ll leave it all running in the background for a day (in between power outages) and see how it goes with the old driver.

8

Although it’s not been a full day yet, the old driver seems to work just fine. I even did the FanControl setup which ended up working better than expected and then also a bit of gaming. With the newer drivers I’m guaranteed to get a freeze within a few hours IF any monitoring software is also running.

The 2x140mm Noctua intake fans that were connected to the GPU headers are now connected with a splitter to my AIO pump header on the motherboard. The AIO header needed a setting changed in the BIOS, don’t recall what it was as I already did it many months ago, and it supports zero RPM (I didn’t know) which means I can still have 2 of my intake fans NOT pull in dust for no reason when the GPU is idling.

The 5600x has a slight undervolt which means it boosts slightly higher and the 1660ti is stock and boosts to over 1900Mhrz out of the box… as long as the temp stays below 60c.

image

image
image1920×417 101 KB

9

I was wondering if turning on hardware scheduling in Windows might make a difference with the newer drivers and possibly solve my freezing issue with newer drivers, but right now I don’t feel like testing that. This has already taken a large chunk of time and I need to get back to my Devember project.

I did submit driver feedback to NVidia…though I’m not holding my breath that they’ll really care much or that the issue will be fixed any time soon.

10

This is a really interesting story, hope you got it fixed now with the older driver.

If you still want to control the front fans as the GPU temperature changes and you have a temperature sensor header on the motherboard you could buy a thermocouple, stick that on the GPU and use it as a source for temperature. That’s how I did it not having any fan headers on the GPU.

11

Don’t even have to do that. You can use FanControl to ramp up fans based on CPU or GPU temps by creating a Mix fan curve. So I can have my 2 fans connected to the AIO header on the motherboard and use FanControl to manage the fans based on GPU temps. I can then have other fans ramp up or down on either GPU or CPU temp. In the image below you can see I have a curve for CPU, a curve for GPU, and then a Mix curve which will ramp up the fans depending on which is hotter.

As per usual, the community is just doing things better :smiley:

image
image813×858 57.5 KB

1 Like
12

I remember seeing that software somewhere else, but I didn’t wanto to tie my setup to a software. That way I won’t need to fiddle with software if I format my system or install Linux. Thanks for the advice!

13

Yea, I would like not to have to use software, but right now this is the only way I can get the setup I want and I can at least backup my config and easily restore after a format with FanControl. It is also the only piece of software I need now.

1 Like
14

FanControl is silly good, gives more control over ones fans than they actually need. +1

1 Like
15

Well… here we are more than 2 weeks later and so far no lockups with the old driver.

FanControl has also been running with zero issues. It’s just doing it’s thing in the background… as it should.

Side note: I have been getting a bluescreen related to memory access… or something… I’m not and expert. Eventually figured out it will constantly happen if I leave the machine to idle for long enough. Some Windows background service is starting up after the machine has been idle for some time which results in a BSOD. How long idle… not sure… 30min+, maybe 1 hour+.

I ended up using WPD to disable and remove a bunch Windows crap and I guess one of the items I disabled/removed was the culprit as I have no more BSOD even after leaving the machine idling for hours. Windows has always had it’s issues, but it’s gotten so much worse with Win10/11.

Anyway, old NVidia driver solved my lock up issues… which eventually resulted in NTFS errors. Amazing how dangerous updates from companies like NVidia and Microsoft are these days. What a time to be alive…

WindowsUpdate

16

This topic was automatically closed 273 days after the last reply. New replies are no longer allowed.