Jump to content
Corsair Community

Black Screening / Video Card output switching off from iCUE


WhiskeyOmega

Recommended Posts

This is still an issue we are working to resolve but to be honest, only a handful of you have ever forwarded me the data. This tells me a few things:

 

1.) Its very system specific to a narrow set of hardware.

 

2.) None of what we are implementing is working which means it could be a compatibility issue with the CPUID SDK (which is what Corsair Service talks to).

 

3.) Some people have told me that they found it wasn't just iCUE triggering this, which complicates things.

 

To complicate things, I can't get any of the hardware you're all using anymore either. So it is something we are working on, but keep in mind I have nothing more than what you all have provided to work with.

 

Hi,

 

I'm experiencing a similar issue with my system haging/freezing for a few seconds while iCue is installed. But no black screen but nvidia driver crash in the windows event log. There seem to be a complete hardware rescan after the crash as al of my usb devices are frozen for a while and windows is asking me about the device that I've connected (head hones or external speakers). See my system specs, I' on a z370 chipset and seem to be having similar issue.

Edited by Daniel_Gagne
Link to comment
Share on other sites

This is still an issue we are working to resolve but to be honest, only a handful of you have ever forwarded me the data. This tells me a few things:

 

1.) Its very system specific to a narrow set of hardware.

 

2.) None of what we are implementing is working which means it could be a compatibility issue with the CPUID SDK (which is what Corsair Service talks to).

 

3.) Some people have told me that they found it wasn't just iCUE triggering this, which complicates things.

 

To complicate things, I can't get any of the hardware you're all using anymore either. So it is something we are working on, but keep in mind I have nothing more than what you all have provided to work with.

 

I replied to your pm from your last version the day you sent me it. It was an invalid link. So I couldnt send you data.

Link to comment
Share on other sites

This is still an issue we are working to resolve

Hi James, thanks to you and your team for your hard work on this. TBH: you've shown a higher level of support than most other companies would and it's great that you're working with the consumers to try to work this out. I can imagine how tough a bug this is to chase.

 

a handful of you have ever forwarded me the data.

I've posted my initial testing results to this thread. I also posted the results of initial troubleshooting after the problem presented itself following the most recent update. I hope they've been at least somewhat helpful since I'm a bit of an outlier running Win7.

 

 

1.) Its very system specific to a narrow set of hardware.

I'd argue that there are a few commonalities I've noticed that make me suspect the effects are wider reaching than that. I see ASUS and AMD hardware included in most, if not all, of the other people I've looked at. There could be a much larger segment that's affected if ASUS/AMD combinations are susceptible to this issue.

 

2.) None of what we are implementing is working which means it could be a compatibility issue with the CPUID SDK (which is what Corsair Service talks to).

You may be correct. I say this because it's always been my gut impression that this is related to hardware polling and therefore the root cause may be on a dependent service rather than your software/implementation.

 

3.) Some people have told me that they found it wasn't just iCUE triggering this, which complicates things.

Running in my current state (CPUIDSDK files renamed, also detailed in my most recent past post), I've encountered *similar* but not identical crashes on a sporadic basis which could further suggest that there are other factors at play that lie beyond your software. Another user mentioned this and I am inclined to agree: there may be a relation with AMD's video card drivers. For instance, I've encountered unrelated crashes of the AMD video driver as well as a recent episode where I encountered strange system hangs in which the video would blank (black screen, no signal to monitor) but the system would appear otherwise functional. This is in direct contrast to the "black screen" crashes where the entire system (down to hardware level) is unresponsive and requires full, hard power cycle. However since the behavior is not identical, there's no way for me to rule out other, completely unrelated, causes for the hangs. And because the behavior stopped as mysteriously as it started, I can't say any of this is related with certainty. It's equally possible an interaction between the video card driver and another piece of software (Java is a likely suspect) could be causing crashes that mimic the issue we've encountered with iCue/CUPIDSDK. I've also observed periodic instances where the graphics card, mouse, and onboard LEDs seem to "freak out" (there have been various behaviors here). Further making things potentially difficult: I'm following a "least invasive" approach to testing since my aim is to keep this system functional being it's my "daily driver". I would be happy to perform additional tests if you think it would be helpful (within reason). I've done a lot of past work in beta testing/troubleshooting so if my skills can be of any help to you, I'd be happy to try.

 

 

To complicate things, I can't get any of the hardware you're all using anymore either. So it is something we are working on, but keep in mind I have nothing more than what you all have provided to work with.

With the exception of my graphics card (and case and HDDs, which I doubt are related), all of my current build was recently purchased (within the past 6 months). I believe I listed the ages of all my hardware in my initial post to this thread as well. I would think you should be able to build a decent replica of at least my system and possibly a few others. Installation of Win7 on this platform was non-standard, but executed properly and is otherwise functioning normally with basically stock configuration (all recent drivers and current updates installed). Although, since this seems to be independent of what OS is running, I would think installing Win10 on this platform may result in the same behavior since I think it's safe to say hardware is definitely a factor here.

 

I'll keep watching this thread loosely for updates. Feel free to reach out to me directly if I can be of service.

Link to comment
Share on other sites

As if to make me look stupid, my PC finally did encounter a "black screen" crash this morning with all the same symptoms. I have not changed anything from the previous config and wasn't doing anything unusual (web browsing), so this seems to point even more to the issue being present independent of CPUIDSDK (or perhaps is greatly exacerbated/accelerated by it).

 

The only light I can shed from my observations seems to be a commonality between unexpected behavior I've noticed with the various HW monitoring software when I wait between starting up and logging in. In short: if I log into the PC right away, the system starts up and will run without incident. If there's a length of time between the login screen appearing and me actually logging into an account, I have noticed a tendency for one or more the following happening after a seemingly random but short period of use:

- Motherboard LED indicators will show high temp (falsely)

- ASUS monitoring software won't be able to read CPU temp

- Mouse LED will go out/display high temp (falsely)

- GPU card fans will go into high speed mode (falsely)

- System hangs

- System crashes (black screen but resettable)

- "Black screen crash" (requires hard power off; has only happened once so far)

This morning I started the PC but didn't log in for several minutes.

 

I'm leaning more and more towards some odd conflict with AMD's HW or SW since it seems like the video card is what's not playing nice. This is only a gut feeling, but sometimes they're spot on.

 

Sorry to muddy the waters further, but wanted to make sure I'd furnished the most accurate information I can.

Link to comment
Share on other sites

Note too, folks ... renaming CPUIDSDK.dll to something different seems to alleviate, if not resolve, the issue. The downside is that you lose system monitoring in iCue.

 

I had uninstalled Icue Completely and my PC had been running for apx 3 weeks straight.

 

Re-Installed and tried the above within 10 minutes of iCue running after rebooting, I had the black screens again.

Link to comment
Share on other sites

Hey guys. I just finished my testing today and sent Corsair James the results by PM with details. In case anyone is interested, here are the summarized results I had on my system over the past couple of weeks or so.

 

Test #1: Black screen DOES appear.

Test #2: Black screen DOES NOT appear.

Test #3: Black screen DOES appear.

Test #4: Black screen DOES NOT appear.

Test #5: Black screen DOES appear.

 

Hopefully Corsair finds this useful and figures this thing out.

 

Until this is fixed, I'm just going to run iCUE like normal, but I'm going to leave the "Corsair Service" service disabled. Sure, I won't have temp monitoring etc. but at least I'll have my G key macros working which is the most important thing I use iCUE for anyway.

Link to comment
Share on other sites

  • 2 weeks later...

While I haven't been able to run the tests, I do have some feedback that may help pinpoint the problem. The last several months I've left iCUE installed on my system, but have prevented it from running, including (most importantly) the Corsair service. I've found that even with iCUE closed, if the Corsair service is running, the crashing problem will occur.

 

Unfortunately, while I would like to help out more, I can't risk these crashes happening on my PC as I use it for streaming, and it tends to be more likely to crash when I'm live.

Link to comment
Share on other sites

Corsair K70 MK.2 Low Profile here. Purchased just last week Monday evening. BSOD on my Win10 64-bit machine about twice a day untill I did #2 of the 5 tests. PC been running fine so far, and iCUE is running in my system tray.

Of course, the downside is that whatever the Service had controlled is not working.

Link to comment
Share on other sites

I'm going to be upgrading my main system to a Z390 mobo with a 9900K shortly. Has anyone heard of this issue happening on those systems? I'm going to load iCUE with everything running normally when I set up the new system, but at the first sign of a black screen issue I'll be turning off the Corsair Service and renaming the service-related files. I don't want any crashing on my fresh new system!
Link to comment
Share on other sites

Renaming only the CPUIDSDK dll has, apparently, been effective for a number of users. This minimizes the lost functionality.

 

As for z390 - I've not seen or heard of this issue with z390. It's pretty specific to one platform in most cases. Plenty of folks are also running this on Z170, Z270 and Z370 without this issue.

Link to comment
Share on other sites

  • 2 weeks later...

Hi All

I have actually just RMA'd my graphics card for this issue. The replacement I got has the same issue. I finally noticed that iCUE was crashing every time (i managed to get it into a reproducible state after weeks of testing)

 

To reproduce it easily, first turn TDR off, this is the windows feature that recovers the graphics card. You'll notice that when it crashes everything just stops, usually for a few seconds, sometimes it dies completely.

 

TdrLevel - Specifies the initial level of recovery. The default is to recover on timeout, which is represented by value 3.

KeyPath : HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\GraphicsDrivers

KeyValue : TdrLevel

ValueType : REG_DWORD

ValueData : 0 - Detection disabled OR 3 - Recover on timeout

 

The test that I could run to get the issue to happen every time was BurnInTest set to CPU maximum heat, and cranked up to 100% CPU usage, whilst having iCUE running and HWiNFO64. Now when you open HWiNFO64 and go into sensors, it prompts me to monitor the ASUS motherboard setting or not, which I ignored for the time being and monitored everything.

 

This would crash, instantly. Many many times, with a small 1-5 second gap then 1-5 seconds of crashing.

 

I realised iCUE was crashing, so I googled, found this reddit post:

 

Which led me here.

 

Now, through further investigation and a lot of reading, I'm pretty confident saying that the issue is iCUE monitoring the ASUS EC, and then running a program like HWiINFO64 will trigger the crash instantly. Whilst after all my testing, I don't believe all these steps are required, they are the ones I took in the process. As per below I'm pretty sure you can force the crash quite easily.

 

I don't believe it is iCUE's fault at all, but as it is an application which we need for all our corsair products to function correctly, many people have it running ALL the time. If for any reason another tool read's the ASUS EC at the same time as iCUE, it will kill the display driver.

 

My PC info:

Intel Core i7-7740X Intel X299 (Kaby Lake)

ASUS ROG STRIX X299-E GAMING

GIGABYTE RTX 2080 GAMING OC

4x 8 GB PC4-25600 DDR4 SDRAM Corsair CMK16GX4M2B3200C16

Link to comment
Share on other sites

Just thought I'd add that since replacing my system with a Z390 Asus motherboard, I have not had any issues with screen freezes or going black. It only happened on my old X79 Asus board.

 

I know others have mentioned that they have not seen this on Z390 boards either, but just thought I'd throw in my 2 cents since I have been running the full install of iCUE for about one week now with no problems.

Link to comment
Share on other sites

I spent about year trying to fix this until I found this thread. I just keep the app from launching, I don't keep my computer as long as I used but when I do run it for a few days without ICUE or LINK running I don't get blackscreens anymore. I know this will go away when I upgrade my X79 build but abandoning a whole chipset because Intel did years ago isn't right.

 

I guess i'm not upset Corsair anymore knowing that it's only affecting the X79 chipset (and ASUS but I don't have ASUS anything) and a upgrade will fix it, at least for me. I like the fact that devs are actually responding to our complaints unlike other companies that have a forum for users to help other users instead of the devs owning it fixing it on their level.

 

Knowing that it's a software based problem and not my components takes a load off knowing that I can hold out until Zen 2 releases for my upgrade.

Link to comment
Share on other sites

I spent about year trying to fix this until I found this thread. I just keep the app from launching, I don't keep my computer as long as I used but when I do run it for a few days without ICUE or LINK running I don't get blackscreens anymore. I know this will go away when I upgrade my X79 build but abandoning a whole chipset because Intel did years ago isn't right.

 

I guess i'm not upset Corsair anymore knowing that it's only affecting the X79 chipset (and ASUS but I don't have ASUS anything) and a upgrade will fix it, at least for me. I like the fact that devs are actually responding to our complaints unlike other companies that have a forum for users to help other users instead of the devs owning it fixing it on their level.

 

Knowing that it's a software based problem and not my components takes a load off knowing that I can hold out until Zen 2 releases for my upgrade.

 

Keep in mind that - at least in most cases - renaming CPUIDSDK.dll to something else does resolve the issue. You'll need to shutdown iCUE and stop the service in order to do this AND you'll have to do it with every update.

Link to comment
Share on other sites

Hi DevBiker,

 

I tried that with ICUE, CPUID ran even though the files were renamed. When I renamed the files when LINK was installed the "fix" worked.

 

That's why I was keep it from running at boot. I don't mind keeping it off as long I know that my computer isn't going to black screen every time something is put under load.

 

Thanks for the reminder for renaming though.

Link to comment
Share on other sites

I'm having issues with iCUE as well, I even reinstalled windows several times till I figured out it was iCUE causing my blue screens.

 

I got rid of my mouse and replaced it with a different brand as I could not get used to it anyway (Corsair Scimitar pro rgb), but am still using the keyboard. (k55).

 

After removing iCUE, no more blue screens.

 

Is there a way to use the macro keys on my K55 keyboard WITHOUT using the iCUE software, or will I need to return this one as well?

Link to comment
Share on other sites

though i didn't have black screen or crashing issues

CLink4 was causing some extreme dpc latency

renaming just the cpuidsdk.dll has fixed this issue for me thank you:biggrin:

i use aida64 for temp monitoring (doesn't cause the latency issue) only need link for my cooling profiles

Link to comment
Share on other sites

  • 2 weeks later...

Had this issue on a Strix 370-F, and a Crosshair VII Wifi, both with a 2700X. Came from a 4690K and Z197(?) Krait which couldn't care less about iCue.

 

Just to clarify, do we need to change the name of both CPUIDSDK.dll AND CPUIDSDKx64.dll (or something like that...the 64-bit version of the .dll?) I did to be safe, but wasn't sure if it was necessary.

 

When it's not the holidays and I have time to do some more digging I will try to run the tests requested by Corsair and see what happens. Definitely one heck of a buzzkill though. Really had me rethinking my Ryzen build for the last three weeks or so.

 

...'Course, that being said, if I did end up going Intel, are there Z370/Z390 boards having issues as well? Would kinda defeat the purpose of doing so obviously...

Link to comment
Share on other sites

  • 2 weeks later...

Just wanted to sound off here this is clearly still an issue - I found this thread from this recently archived reddit thread about this issue below. I was gifted the Corsair Wired Void Pro RGB headphones for Christmas. I installed the iCUE software only to be greeted by my PC randomly locking up and eventually recovering after about 30 or so seconds. The PC's screen would go black, sound would play for a second or so, then all would be non-responsive with a lasting black screen. Eventually the pc would recover and the monitor would display my desktop/apps again. At first I thought this was a driver crash, but after sleuthing around I discovered the below reddit thread and this thread and knew it was iCUE. Like many, I too have an ASUS motherboard; it seems to be related to folks who have an ASUS board and possibly use the AURA software to control lighting on the board (I do).

 

I immediately uninstalled iCUE and installed the older CUE software and my system has been stable ever since (been stable going on 4 days now, I was getting PC lock ups ever few hours prior).

 

I don't know if corsair staff is still monitoring this thread (a brief look through the many pages of this and it looks like the last official corsair staff person I saw post was back in October unless I missed a newer post) but I wanted to add not only is iCUE causing havoc on mine, and many other peoples systems, it also is extremely laggy/jittery when moving it around on a high refresh rate monitor (I have a 144hz display and it was like looking at a slide show when moving the app around). This is the only piece of software I've used that isn't smooth moving on the display (a minor gripe I know compared to crashing/freezing systems, but still really annoying).

 

Corsair staff if you are still monitoring this, please give an update on your efforts to fix this issue.

 

Reddit Thread:

Link to comment
Share on other sites

Just had this issue happen on my PC, here are details from the event log:

 

Faulting application name: iCUE.exe, version: 3.10.125.0, time stamp: 0x5bf8698e

Faulting module name: libGLESv2.dll, version: 0.0.0.0, time stamp: 0x5a5da43b

Exception code: 0xc0000005

Fault offset: 0x000da111

Faulting process id: 0x2170

Faulting application start time: 0x01d4a429fcd93d3a

Faulting application path: C:\Program Files (x86)\Corsair\CORSAIR iCUE Software\iCUE.exe

Faulting module path: C:\Program Files (x86)\Corsair\CORSAIR iCUE Software\libGLESv2.dll

Report Id: 437e189f-2afc-49b1-a9b6-3acb58f18548

Faulting package full name:

Faulting package-relative application ID:

 

Fault bucket 1383692415130177826, type 1

Event Name: APPCRASH

Response: Not available

Cab Id: 0

 

Problem signature:

P1: iCUE.exe

P2: 3.10.125.0

P3: 5bf8698e

P4: libGLESv2.dll

P5: 0.0.0.0

P6: 5a5da43b

P7: c0000005

P8: 000da111

P9:

P10:

 

Attached files:

\\?\C:\ProgramData\Microsoft\Windows\WER\Temp\WERB4A7.tmp.dmp

\\?\C:\ProgramData\Microsoft\Windows\WER\Temp\WERB5A2.tmp.WERInternalMetadata.xml

\\?\C:\ProgramData\Microsoft\Windows\WER\Temp\WERB5B3.tmp.xml

\\?\C:\ProgramData\Microsoft\Windows\WER\Temp\WERB5B1.tmp.csv

\\?\C:\ProgramData\Microsoft\Windows\WER\Temp\WERB5C1.tmp.txt

 

These files may be available here:

\\?\C:\ProgramData\Microsoft\Windows\WER\ReportArchive\AppCrash_iCUE.exe_583f6d93fabcfa925ebf5ec2c8a5329b71cca2c_cce0e3ef_4fe2b860

 

Analysis symbol:

Rechecking for solution: 0

Report Id: 437e189f-2afc-49b1-a9b6-3acb58f18548

Report Status: 268435456

Hashed bucket: df85e9fb4d75f3f0f333dceba23c9122

Cab Guid: 0

 

Faulting application name: iCUE.exe, version: 3.10.125.0, time stamp: 0x5bf8698e

Faulting module name: libGLESv2.dll, version: 0.0.0.0, time stamp: 0x5a5da43b

Exception code: 0xc000041d

Fault offset: 0x000da111

Faulting process id: 0x2170

Faulting application start time: 0x01d4a429fcd93d3a

Faulting application path: C:\Program Files (x86)\Corsair\CORSAIR iCUE Software\iCUE.exe

Faulting module path: C:\Program Files (x86)\Corsair\CORSAIR iCUE Software\libGLESv2.dll

Report Id: a20e0953-d7b5-49b1-8436-18554686b0ba

Faulting package full name:

Faulting package-relative application ID:

 

Fault bucket 2236802420129695949, type 1

Event Name: APPCRASH

Response: Not available

Cab Id: 0

 

Problem signature:

P1: iCUE.exe

P2: 3.10.125.0

P3: 5bf8698e

P4: libGLESv2.dll

P5: 0.0.0.0

P6: 5a5da43b

P7: c000041d

P8: 000da111

P9:

P10:

 

Attached files:

\\?\C:\ProgramData\Microsoft\Windows\WER\Temp\WERB93B.tmp.dmp

\\?\C:\ProgramData\Microsoft\Windows\WER\Temp\WERBA26.tmp.WERInternalMetadata.xml

\\?\C:\ProgramData\Microsoft\Windows\WER\Temp\WERBA37.tmp.xml

\\?\C:\ProgramData\Microsoft\Windows\WER\Temp\WERBA37.tmp.csv

\\?\C:\ProgramData\Microsoft\Windows\WER\Temp\WERBA48.tmp.txt

 

These files may be available here:

\\?\C:\ProgramData\Microsoft\Windows\WER\ReportArchive\AppCrash_iCUE.exe_1c57db45d8072e422499226aa9de5c1eb65c113_cce0e3ef_56e2bbfa

 

Analysis symbol:

Rechecking for solution: 0

Report Id: a20e0953-d7b5-49b1-8436-18554686b0ba

Report Status: 268435456

Hashed bucket: a5690e3df87270d0ef0ab7f3349088cd

Cab Guid: 0

Link to comment
Share on other sites

Just checking in to note that I have the same problem with iCue. Any time iCue has run on my current rig, software that utilizes Iray for rendering becomes unstable to the point of being unusable. When I try a render, I get the exact same black screen behavior and then am greeted by an error message when the screen comes back.

 

I've found this issue to be 100% repeatable. If I don't let iCue run at boot, then I can render all day long without any issues. The very first render attempt after loading iCue ends with the black screen cycle.

 

Since this rig was built for rendering and Corsair won't let me render on it, I don't think I can justify having Corsair controls in it much longer. What a disappointment.

Link to comment
Share on other sites

×
×
  • Create New...