Yeah... I know. It isn't the drivers, I'm fairly certain... (I stay up to date on those). Could be cable, could be monitor. But this only has happened on this new Alienware box, not on my previous PC.
I might try re-seating the GPU and switching out cables. I am good at troubleshooting, I just thought before I started, I would ask the community if they'd seen anything similar. Thanks.
Could be a number of things, usually an issue with the gpu drivers, the connection from gpu to monitor, or the monitor. I just had this issue with a work laptop I got in December, but since it was business inventory I just swapped it out with a different model.
Addendum: I ram a couple of stress tests for about 10 minutes. Prime95 and FurMark. Seems fine.
I know you're supposed to run these for hours, but if there was something drastically wrong with power, or cooling, a few minutes in either of these would probably show it. But the PC has no problem... all the fans are running, the temps are under control, and it didn't trigger any black screens.
Ok, you managed to get the system through the hoops but I presume this is basic settings? Sometimes it takes a little push near to the edge for problems to expose themselves. Here's my suggestion:
1. Wipe the system and do a clean reinstall of Win10.
2. After Win10 downloads all the updates needed, check if Alienware update has anything especially BIOS. If there's a BIOS update, do that. Otherwise Dell technical support is going to ask you to do so anyways (part of the standard scripted response).
3. Once that is all done, see if the random 2-sec black screen appears. If yes, then go to NVIDIA's website and download the latest drivers for your card. This is where some of us here in the community may differ in opinion. One camp like the one I'm in, will take NVIDIA site posted drivers. Another camp will ask you to stick to DCH (OEM signed) drivers. If you are already on the latest DCH drivers per Dell update website, then it's time to look at the NVIDIA site ones, agree?
Disconnect the R8 from the Internet and use Add/ Remove program to uninstall the NVIDIA drivers and then install the one you downloaded. Restart, and turn back on the Internet connection.
4. If problem still persists, rerun the Prime95 and Furmark tests. 10 minutes is fine but I'd throw in 3Dmark Time Spy stability test just to be safe. Another suggestion is to use Heaven because the plethora of scenes forces your GPU to render and put most if not all of its units to exercise. My hunch is that if there's a very specific hardware issue with your card, you'll see a significant dip (think single digit fps) for a scene.
5. If none of the tests above have problems, then up the OC and Boost on the CPU and GPU. Assuming you're familiar with OC of these parts, I won't go into the details. Rerun those tests. Only thing to note is it's also good to split the OC of GPU core clock and memory clock separately so that if there's a bad GDDR on board, you'll know right away.
Let us know how it goes. If you have another PC that supports the 2080 Super, why not transplant it there to see if the issue is with the GPU or not?
OH one other thing: I did run Prime95 and FurMark (stress mode) for an hour or so, just to see if something was obviously flaky... but had no problems. If something is wigging out on its own for 2 seconds systems-wise, I would expect a stress test would fail rather quickly. Thanks.
I hear ya man... after 25 years in the IT industry, troubleshooting just comes as second nature. The Win10 install is fine (hasn't been on disk long enough for me to screw it up too badly ) ... I try to spend as little time possible reinstalling Windows, troubleshooting components, etc. these days. Before I did that, was gonna ask if anyone else had seen this behavior (I imagine an awful lot of R8s got sent out in the last month, due to the good sale.)
Hmm, I would think that most people would jump on the R9 bandwagon by now but looks like many still appreciate a good deal and the non-Legend design. Anyways, I concur that stress test like what you've run would have caused the system to barf if there's a fundamental issue.
You said that the blank screen doesn't happen to your old system using the same monitor and cable. Perhaps the tolerance of the new card is less when it comes to HDMI/ DP. BTW, are you using HDMI or DP? Variable or fixed refresh? One thing I would check would be to see if the system is still "alive" when the blank-out happens. I'd go something like playing some music or video and see if it freezes together with the blank screen or life just goes on as normal, with only a loss in display.
Try swapping out the HDMI/ DP cable. Assuming you're here in the US, Amazon Basics cable should be ok for testing. Just get a reasonable 3-5 ft length.
1. It isn't the drivers, I'm fairly certain... (I stay up to date on those).
2. The Win10 install is fine (hasn't been on disk long enough for me to screw it up too badly ) ... I try to spend as little time possible reinstalling Windows,
3. The screen goes blank/black, as if the connection between the video card and the monitor is broken for around 2 seconds. Not long enough for my monitor to complain "no signal", but definitely long enough to cause worry.
1. The latest Nvidia drivers are not always trouble-free. Often, it is not the most stable or even recommended for a particular card.
2. You seem to be under the misconception that the Windows, drivers, and Apps that Dell loaded onto your particular machine is somehow infallible ? What gave you the idea that it is perfect (or still perfect)?
3. Sounds like a HDCP issue to me.
I thought @GTS81 posts were rather good myself. What he is trying to do is help you determine if it is hardware or software. That is where you start.
Oh and regarding the preinstalled Dell/Alienware stuff... that's the kind of thing I was asking about. Has anyone had any major troubles with it? (Looks like the AWCC is the largest component...)
I usually wipe all that stuff and install an OS cleanly with prebuilt machines like this, but I was pleasantly surprised that the pre-installed "crapware" wasn't too bad with the R8. But you are right of course, all software has bugs.
Yes it was a good post! There was just a miscommunication; since it's only happened about 3 times and isn't reproducible, I'm not trying to troubleshoot the problem; just wondered if anyone had seen that in recent R8 Cheers.
OK it JUST happened. I went into Event manager, and I noticed this around the exact time that the screen went blank for about a second:
Dell Digital Delivery : System.BadImageFormatException: Could not load file or assembly 'Dell.D3.HSA.Server.dll' or one of its dependencies. The module was expected to contain an assembly manifest. File name: 'Dell.D3.HSA.Server.dll' ---> System.BadImageFormatException: Could not load file or assembly 'Dell.D3.HSA.Server.dll' or one of its dependencies. The module was expected to contain an assembly manifest. File name: 'Dell.D3.HSA.Server.dll'
at System.Reflection.AssemblyName.nGetFileInformation(String s) at System.Reflection.AssemblyName.GetAssemblyName(String assemblyFile) at System.ComponentModel.Composition.Hosting.AssemblyCatalog.LoadAssembly(String codeBase) at System.ComponentModel.Composition.Hosting.AssemblyCatalog..ctor(String codeBase) at Dell.D3.WinSvc.Controller.DependencyInjector.SafeDirectoryScan(AggregateCatalog catalog, String searchPattern) Dell Digital Delivery : BadImageFormatException when loading C:\Program Files (x86)\Alienware Digital Delivery Services\Dell.D3.HSA.Server.dll
I actually had removed it a few days ago, once I decided I didn't need it (mostly good for overclocking and changing the LED lights, which I don't care about...) Then I re-added it when i was messing with the side-panel (I know you know the story on that one!) just to see if the side panel's LED was working.
I reckon I'll remove all Alienware-branded utilities/etc and see if I still get the intermittent blanking. I know you can change the RAM overclocking in UEFI; I wonder about the CPU overclocking? (I'm only using the first overclocking "pre-set", because the PC was advertised as having "out of the box" OC).
Thanks man, will let you guys know if it keeps happening!
EDIT: The bad thing about this is, it's not predictable or reproducible, so making a change may or may not "help"... but I reckon that's a good place to start.
Have you tried control panel, add remove programs, Alienware Digital Delivery Services? That is one program I did not reinstall after clean installing windows on my system.
Alright I removed all Alienware/Dell branded stuff in Apps, beside "Alienware Product Registration" and "Alienware Update for Windows 10" ... we'll see if I get the random blanks anymore. Thanks!
@GTS81 Oh and, when you reinstalled Windows, did you just use the vanilla "Windows 10 USB Installer" thing from Microsoft?
Reason I ask is, on my OLD computer, the reinstall DVD of Windows wasn't even high enough of a patch level to support the ethernet chip on the motherboard; you had to use the Windows installer, then a driver disc to install the driver for the ethernet chip (granted, I bought the PC right after Win10 was released, and the motherboard was very very new...)
With the Alienware, I'm not sure what the default "reset back to factory" option would be, since it doesn't come with a driver disk. I guess you're just supposed to use the "Reinstall Windows for me" tool, but... what if I had put Linux on the darn thing, and wanted to reinstall Windows? Ahhh, PCs are such fun, aren't they?
Mercster
35 Posts
0
February 1st, 2020 07:00
Yeah... I know. It isn't the drivers, I'm fairly certain... (I stay up to date on those). Could be cable, could be monitor. But this only has happened on this new Alienware box, not on my previous PC.
I might try re-seating the GPU and switching out cables. I am good at troubleshooting, I just thought before I started, I would ask the community if they'd seen anything similar. Thanks.
r72019
6 Professor
•
5.3K Posts
1
February 1st, 2020 07:00
Could be a number of things, usually an issue with the gpu drivers, the connection from gpu to monitor, or the monitor. I just had this issue with a work laptop I got in December, but since it was business inventory I just swapped it out with a different model.
Mercster
35 Posts
0
February 1st, 2020 07:00
Addendum: I ram a couple of stress tests for about 10 minutes. Prime95 and FurMark. Seems fine.
I know you're supposed to run these for hours, but if there was something drastically wrong with power, or cooling, a few minutes in either of these would probably show it. But the PC has no problem... all the fans are running, the temps are under control, and it didn't trigger any black screens.
Thanks.
GTS81
2 Intern
•
2.2K Posts
0
February 1st, 2020 12:00
@Mercster :
Ok, you managed to get the system through the hoops but I presume this is basic settings? Sometimes it takes a little push near to the edge for problems to expose themselves. Here's my suggestion:
1. Wipe the system and do a clean reinstall of Win10.
2. After Win10 downloads all the updates needed, check if Alienware update has anything especially BIOS. If there's a BIOS update, do that. Otherwise Dell technical support is going to ask you to do so anyways (part of the standard scripted response).
3. Once that is all done, see if the random 2-sec black screen appears. If yes, then go to NVIDIA's website and download the latest drivers for your card. This is where some of us here in the community may differ in opinion. One camp like the one I'm in, will take NVIDIA site posted drivers. Another camp will ask you to stick to DCH (OEM signed) drivers. If you are already on the latest DCH drivers per Dell update website, then it's time to look at the NVIDIA site ones, agree?
Disconnect the R8 from the Internet and use Add/ Remove program to uninstall the NVIDIA drivers and then install the one you downloaded. Restart, and turn back on the Internet connection.
4. If problem still persists, rerun the Prime95 and Furmark tests. 10 minutes is fine but I'd throw in 3Dmark Time Spy stability test just to be safe. Another suggestion is to use Heaven because the plethora of scenes forces your GPU to render and put most if not all of its units to exercise. My hunch is that if there's a very specific hardware issue with your card, you'll see a significant dip (think single digit fps) for a scene.
5. If none of the tests above have problems, then up the OC and Boost on the CPU and GPU. Assuming you're familiar with OC of these parts, I won't go into the details. Rerun those tests. Only thing to note is it's also good to split the OC of GPU core clock and memory clock separately so that if there's a bad GDDR on board, you'll know right away.
Let us know how it goes. If you have another PC that supports the 2080 Super, why not transplant it there to see if the issue is with the GPU or not?
Mercster
35 Posts
0
February 1st, 2020 12:00
OH one other thing: I did run Prime95 and FurMark (stress mode) for an hour or so, just to see if something was obviously flaky... but had no problems. If something is wigging out on its own for 2 seconds systems-wise, I would expect a stress test would fail rather quickly. Thanks.
Mercster
35 Posts
0
February 1st, 2020 12:00
I hear ya man... after 25 years in the IT industry, troubleshooting just comes as second nature. The Win10 install is fine (hasn't been on disk long enough for me to screw it up too badly ) ... I try to spend as little time possible reinstalling Windows, troubleshooting components, etc. these days. Before I did that, was gonna ask if anyone else had seen this behavior (I imagine an awful lot of R8s got sent out in the last month, due to the good sale.)
Thanks bud.
GTS81
2 Intern
•
2.2K Posts
0
February 1st, 2020 13:00
Hmm, I would think that most people would jump on the R9 bandwagon by now but looks like many still appreciate a good deal and the non-Legend design. Anyways, I concur that stress test like what you've run would have caused the system to barf if there's a fundamental issue.
You said that the blank screen doesn't happen to your old system using the same monitor and cable. Perhaps the tolerance of the new card is less when it comes to HDMI/ DP. BTW, are you using HDMI or DP? Variable or fixed refresh? One thing I would check would be to see if the system is still "alive" when the blank-out happens. I'd go something like playing some music or video and see if it freezes together with the blank screen or life just goes on as normal, with only a loss in display.
Try swapping out the HDMI/ DP cable. Assuming you're here in the US, Amazon Basics cable should be ok for testing. Just get a reasonable 3-5 ft length.
Good luck!
Tesla1856
8 Wizard
•
17.4K Posts
0
February 1st, 2020 19:00
1. The latest Nvidia drivers are not always trouble-free. Often, it is not the most stable or even recommended for a particular card.
2. You seem to be under the misconception that the Windows, drivers, and Apps that Dell loaded onto your particular machine is somehow infallible ? What gave you the idea that it is perfect (or still perfect)?
3. Sounds like a HDCP issue to me.
I thought @GTS81 posts were rather good myself. What he is trying to do is help you determine if it is hardware or software. That is where you start.
Mercster
35 Posts
0
February 2nd, 2020 09:00
Oh and regarding the preinstalled Dell/Alienware stuff... that's the kind of thing I was asking about. Has anyone had any major troubles with it? (Looks like the AWCC is the largest component...)
I usually wipe all that stuff and install an OS cleanly with prebuilt machines like this, but I was pleasantly surprised that the pre-installed "crapware" wasn't too bad with the R8. But you are right of course, all software has bugs.
Mercster
35 Posts
0
February 2nd, 2020 09:00
Yes it was a good post! There was just a miscommunication; since it's only happened about 3 times and isn't reproducible, I'm not trying to troubleshoot the problem; just wondered if anyone had seen that in recent R8 Cheers.
Mercster
35 Posts
0
February 2nd, 2020 10:00
OK it JUST happened. I went into Event manager, and I noticed this around the exact time that the screen went blank for about a second:
Dell Digital Delivery : System.BadImageFormatException: Could not load file or assembly 'Dell.D3.HSA.Server.dll' or one of its dependencies. The module was expected to contain an assembly manifest.
File name: 'Dell.D3.HSA.Server.dll' ---> System.BadImageFormatException: Could not load file or assembly 'Dell.D3.HSA.Server.dll' or one of its dependencies. The module was expected to contain an assembly manifest.
File name: 'Dell.D3.HSA.Server.dll'
at System.Reflection.AssemblyName.nGetFileInformation(String s)
at System.Reflection.AssemblyName.GetAssemblyName(String assemblyFile)
at System.ComponentModel.Composition.Hosting.AssemblyCatalog.LoadAssembly(String codeBase)
at System.ComponentModel.Composition.Hosting.AssemblyCatalog..ctor(String codeBase)
at Dell.D3.WinSvc.Controller.DependencyInjector.SafeDirectoryScan(AggregateCatalog catalog, String searchPattern)
Dell Digital Delivery : BadImageFormatException when loading C:\Program Files (x86)\Alienware Digital Delivery Services\Dell.D3.HSA.Server.dll
Ring any bells?
Mercster
35 Posts
0
February 2nd, 2020 11:00
I actually had removed it a few days ago, once I decided I didn't need it (mostly good for overclocking and changing the LED lights, which I don't care about...) Then I re-added it when i was messing with the side-panel (I know you know the story on that one!) just to see if the side panel's LED was working.
I reckon I'll remove all Alienware-branded utilities/etc and see if I still get the intermittent blanking. I know you can change the RAM overclocking in UEFI; I wonder about the CPU overclocking? (I'm only using the first overclocking "pre-set", because the PC was advertised as having "out of the box" OC).
Thanks man, will let you guys know if it keeps happening!
EDIT: The bad thing about this is, it's not predictable or reproducible, so making a change may or may not "help"... but I reckon that's a good place to start.
r72019
6 Professor
•
5.3K Posts
0
February 2nd, 2020 11:00
Have you tried control panel, add remove programs, Alienware Digital Delivery Services? That is one program I did not reinstall after clean installing windows on my system.
Mercster
35 Posts
0
February 2nd, 2020 11:00
Alright I removed all Alienware/Dell branded stuff in Apps, beside "Alienware Product Registration" and "Alienware Update for Windows 10" ... we'll see if I get the random blanks anymore. Thanks!
Mercster
35 Posts
0
February 2nd, 2020 12:00
@GTS81 Oh and, when you reinstalled Windows, did you just use the vanilla "Windows 10 USB Installer" thing from Microsoft?
Reason I ask is, on my OLD computer, the reinstall DVD of Windows wasn't even high enough of a patch level to support the ethernet chip on the motherboard; you had to use the Windows installer, then a driver disc to install the driver for the ethernet chip (granted, I bought the PC right after Win10 was released, and the motherboard was very very new...)
With the Alienware, I'm not sure what the default "reset back to factory" option would be, since it doesn't come with a driver disk. I guess you're just supposed to use the "Reinstall Windows for me" tool, but... what if I had put Linux on the darn thing, and wanted to reinstall Windows? Ahhh, PCs are such fun, aren't they?
Thanks again.