Other 3D software do not run in window and not using OpenGL. Do you have BSODs on multiple different cards? Within one desktop or few? This might make a big difference if BSOD is caused by GPU driver or firmware. You actually give me a good idea. You can try to run different Unigine benchmarks in OpenGL mode? It's will be a bit near to PA than other benchmarks. Do not overreact just because some post on forum, it's looks silly. PS: I can also notice that you might try to install latest drivers for your soundcard and headset, just to be sure it's not affect system.
Different AMD based cards on the same desktop. I tried multiple versions, no noticeable difference. My BSODs are so rare I cant really measure the time between them and see if a driver improved it a bit. I have a "blank" image with no GPU driver I can roll back for different versions . Ill try so. Better yet, Ill unplug them. No driver loaded - no chance for bad things to happen.
Then this really might be related to some other of your hardware. You never know what cause such problems and I doubt it's will be easy to track it down. You need at least one sound device, otherwise PA will crash on start.
May I ask it the other way round: Is there anything special PA does over other games and applications? Any exotic OpCode-KungFu? Im tempted to try it on a completely different rig too. Ha, nice one, thanks for warning me .
As I said before: PA use OpenGL for all platforms and use virtual textures which is very rare in games. ATI/AMD drivers never had perfect OpenGL implementation, this why people who using professional software (Maya, many CADs and simulation tools) prefer Nvidia GPUs. PA UI is quite unique because it's based on Chromium browser (Google Chrome), but I doubt it's might be reason of BSOD. PA also doesn't use modal fullscreen, that might trigger some Windows/driver bugs too.
Uber wouldn't debug/disassemble/fix missbehaving kernelspace binaries because that is outside of their remit. Obtuse post in an attempt to blame Uber has been seen before... I do tech for HoN and have done since closed beta and the number of people who blame S2 for their machine BSoD is just disgusting whats worse is they refuse to even entertain the prospect something is wrong there end and there is nothing S2 can do This isn't Uber or PA's fault, PA doesn't bust the kernel. THIS is the mindset that needs to change. BSoD are Kernel panics, PA doesn't touch the kernel. It requests things that only kernelspace access can do (access file, allocate memory, draw rasters...) BUT this is via a driver-level API and thus any panics is downto kernel/driver/hardware. The loose sarcasm isn't helping with the acceptance that the perpetuation of BSoD is always software (it might be, not in this instance). PA only works on Vista/7/8 for windows so immediately you have UAC. If kernel-space access was required by PA then it would have to be run as admin. So what causes this one: its a kernel panic and mem related. Its either dodgy memory (RAM or VRAM) or some driver did something dodgy. The dmp would give an indication as to what was going on. This will allow you to view dmp files and determine what was going on http://www.nirsoft.net/utils/blue_screen_view.html
Did you try running memtest and making sure its not your RAM? Planetary Annihilation uses a boatload of RAM... and I've seen plenty of RAM sticks cause blue screening
Well, thats what I used to extract that bits of information. Regarding RAM errors, memory controllers,....: This machine runs huge panostitching that uses most if not all of its 15GB and 100% load on all cores. I ran various memtests for hours. No other application, most beeing much more memory intensive than PA is ever caused a problem. Thats why I fail to beleive in faulty RAM. Currently Im setting up Unigine, running it in OpenGL for some time. Next few days Ill try it on a different machine too. And this is exactly why I think that spamming info about specific bluescreens can be a waste of time because Uber can do close to nothing about it. All they see is a potentially misbehaving foreign binary. Nothing to be done on that. In my first post I stated something like "its a drivers problem". Whatever application causes a condition to expose this drivers problem triggers the crash without doing anything wrong itself. The point I was trying to make: Its doesnt HAVE to be a hardware problem in each and every case.