djvanyel

I've had this issue on an X470 chipset board with a Ryzen 7, which is not old at all (build is coming up on 1 year old). I've mentioned this fact in previous replies. I fail to see why they're insisting on finding an old X79 board when they could have easily re-created a build similar to mine and most likely re-created the issue. I do buy the fact that it's difficult to trace and understand the number of potential variables, but by now I'd expect they'd have partnered with ASUS to try to narrow down the likely culprits. I also happen to have crashes despite having the DLL's renamed, however they are far less frequent (once a week or so). I find that the time between system start and login is directly related to whether or not the system crashes once login completes. I usually have to be quick about typing in a password or I'll end up with a black screen within minutes of logging in. We can only hope for an answer... but after all this time (and all the unsuccessful updates) I'm not holding my breath.

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.

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. 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. 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. 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. 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. 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.

I can confirm that crashing still occurs under 3.7.99 at this time. My system has had no major changes since my last troubleshooting and reply. I was previously running fine, although I've encountered a couple crashes in the meantime but not with the same symptoms as the "Black Screen" crashing from before. When the Black Screen/hard power off required crashing returned after performing the update, I made some notes during troubleshooting as follows: Updated to latest version 3.7.99 via iCUE. No issues encountered during install. Install completed, no reboot needed. Opened iCUE and saw multiple new sensor inputs available (ASUSTEK and others) instead of my previous state when only the video card sensor was only shown in the dropdown list (all other previous settings were intact; see my original reply). Crashing began soon after completing update during light use with exact same symptoms as before. I followed the steps in the previous testing guide above to disable cpuidsdk.dll (stop Corsair Service, rename file), rebooted. Crash occurred. Used same steps to disable cupidsdk64.dll, rebooted. Crash occurred. Attempted to stop Corsair service on next startup, but experienced a spontaneous reboot after opening the Services control panel (no blue screen, just a full reset). After the reboot, I was able to load the Services CP, stopped the Corsair service, and renamed cpuidwin7sdk.dll. Rebooted. System seems stable. I checked over the latest service logs visually via Notepad. No error entries were shown, just many "INFO: DevicesMonitor" entries. Quick scan of the logs did show a few entries where a large number of non-displayed characters are inserted before the entry (directly before the time stamp). There wasn't any other direct correlation between the entries. Not sure if this indicates anything but it appears to correspond to periods after the recent crashes occurred (can't be sure). Checked iCUE settings, now appear as before (only GPU temperature sensor monitoring available). I currently seem stable and running as I was as of my last reply. If anything else of note occurs in the near future, I'll edit this reply as appropriate. Hope this helps.

So far it's only happened the one time. All the crashes seemed to be timed within 5 mins of system startup. This return of the issue could just be a one-off but I'm monitoring to see if it happens again. So far things are back to being stable.

After running for nearly a week under varying conditions with no problems, I encountered a return of the same black-screen crash behavior from before. This is the first instance, and occurred just after completing login to Windows. I have made no changes since my last post and no major updates have been received. Maybe cpuidsdk.dll is not the culprit but only a contributing factor. If I can replicate the issue and obtain anything of value in troubleshooting, I'll update more.

Sorry if I'm somewhat tardy to this party, but just encountered this issue last night and decided to take advantage of a new install to do the tests. Disclaimer: I'm aware of the unusual setup and "unsupported" OS, however I thought since I moonlight as a tester, have a different configuration than most, and my install is brand new, my test results might be useful to the devs in narrowing down the problem. I've provided as much detail as possible to that effect. tl;dr: Skipped test 1, tests 3 and 5 resulted in crash, tests 2 and 4 did not. Running with cpuidsdk.dll renamed appears to prevent crashing. The detailed log is as follows: System configuration (all purchased new, retail, ages shown): -ASUS x-470 Prime (1 week), BIOS version 4011 (most current, as received) -AMD Ryzen7 2700 (1 week) -16GB (2x8GB) Corsair Vengance LPX CMK16GX4M2B3000C15 (1 week) -Saphire Radeon 4870HD (unknown age, running latest stock driver, not Catalyst) -Corsair CX750M PSU (2 weeks) -Corsair Glaive RGB (1 day) -Win7 x64 (stable fresh installation, approximately 1 week old) -System tested under gaming conditions less than 24 hours before new mouse installation with no issues. -All updates installed and drivers are all updated to most current available. -iCue downloaded upon installation of mouse according to in-box instructions. Installed successfully as expected. During initial configuration, I found that the Temperature setting, when monitoring the CPU, would always indicate red, despite the CPU being between 39 and 45. Motherboard LED effects set to monitor CPU temperature indicating below 45 consistently with fans running at system-goverened reduced speed. When set to monitor the graphics card temperature, LED effects indication appeared to be correct and would change if setpoints were altered. Crashes began after completing initial configuration under admin and sub-admin accounts. Initial crash caused black screen (video out; monitor reports no signal); keyboard, mouse, and USB network adapter all disconnected (no power). Motherboard completely unresponsive to all buttons (power and reset). Error indicators on graphics card all lit (red x3, experience suggests this indicates a general system fault). Had to shut off system via PSU switch. Reboot completed normally. Second crash caused disconnection of mouse and black screen with same conditions as before. Keyboard disconnected upon attempting input. USB network adapter appeared to be connected electrically but with no system access evident (steady status LED). Consecutive crashes occurred in similar fashion. All require forcible power down via PSU switch. Gut feeling says it's a hardware polling issue. Test 1: Did not perform. All following tests done with pre-existing monitoring/control applications installed and unmodified from previous known good, standard settings. Test 2: Loaded Windows Services. Noticed that mouse LED lighting turned off but came back on. Attempted to stop service. Initially was unable to, Windows error indicates service did not return an error. Second attempt to stop service was successful. Disabled service, rebooted. Mouse lights were initially on during startup, but went out upon completion of system start. Confirmed neither Corsair.Service.EXE nor Corsair.Service.DisplayAdapter.EXE were running. iCue running. Brough iCue to foreground via systray icon. Encountered error regarding profile conflict. Used "Save" option to recover profile, mouse LED effects still out. Attempted changing profile, found that Gradient, Solid, and Temperature no longer function; others work normally. Temperature no longer has selection menu for monitored sensor. Found warning in iCue: "Failed to start CLink service". Attempted to restart via button, warning returned after short period. Functionality of iCue otherwise seems normal; was able to re-set default DPI setting. Did not encounter crash during use (approx 20 min). Test 3: Performed as instructed. Completed reboot without incident. Verified conditions were met. Noted mouse LED lighting effects now working (default color). Crash occurred after approx. 5 min. Test 4: Performed as instructed. Reboot completed without incident. Verified conditions were met. Noted LED lighting effects working (default color). Used system monitoring application (ASUS AISuite) without issue. System stable after over 40 minutes of use (idle, adding to this log, and read the test reports in the thread), no issues. Test 5: Performed as instructed. Reboot completed without incident. Verified conditions were met. Mouse LED lighting effects unchanged. Crash occurred in under 5 mins with virtually no usage. Rebooted system, reverted to Test 4 state. Corsair services running as before. System run for approx 30 mins with no issue. Loaded iCue, again encountered Profiles Conflict error. Used "Save" option without issue. LED light effects went dark. Set new color profile (solid color, different from default) without issue. Found Solid and Gradient not working but Temperature now does. Selection menu for Temperature source now only shows graphics card but effects work as expected. Chose setting, changed midrange to 45. Left iCue running minimized. No change after approx. 20 mins. Ran system for over 40 mins, no issues. **iCue side note: discovered that if any of the button assignments under "Actions" are highlighted, the highlighted button does not work unless you click out (clear highlights). This persists even if the Actions tab is not selected and iCue is minimized to the systray.** Re-enabled iCue "Start on system startup". Noticed that the "Enable SDK" was enabled where it may not have been previously. Rebooted. Reboot completed without issue. Corsair services and iCue all running. Loaded GUI via systray icon and again encountered "Profiles conflict" error. Using "Cancel" option and going into Settings tab caused iCue to hang. Terminated iCue forcibly. Restarted iCue, used "Save" option, re-selected DPI setting, iCue working same as after reverting from Test 5 to Test 4 state. System ran normally for rest of night. No crashes experienced. Currently running with cpuidsdk.dll renamed to cpuidsdk.dll.bak with no apparent issues. Hope this is helpful.