Interesting thought about overheating! Something like hwinfo might let you see if this is the case or not, just by letting you see the relevant temperatures.
(There's no way to re-init the graphics short of restarting the game; this just isn't a failure case that can really be accounted for. I'm not even sure what it'd *take* to get it back running without a relaunch or if that's even possible.)
Thanks for the info regarding what can/cannot be done regarding the engine.
googling hwinfo seems to be some sort of diagnostic toolkit, I will look into that and other options. I am a bit leery of freeware software, as all too often it is masquerading malware.
Doing some deeeep dives, I have encountered other cases where igfxn crashing and restarting has been causing problems in other software/games. I came upon one *possible* solution, but i won't explicitly say what it involves, as it involves tweaking registry. Due to the ease that someone can royally destroy their computers while fiddling with that, I am going to suggest if people want to know what it is, to message me here and I can send them the details privately. Unless consensus is otherwise?
essentially what the *fix* does is change some parameters regarding when windows forces hardware to restart hardware components (in this case, GPU) when it fails to respond after a set time (in the seconds). The name for this, from what I can tell, is something called
TdrDelay. One avenue using this, is to extend the delay, to give the GPU which has stopped responding (because it is crunching something particularly hard) more time attempt to complete what it is doing, before windows restarts it (which causes the crash).
So far, I have tried this, and I couldn't honestly say if it has improved things or not. The only real concrete evidence that it is doing anything at all is that the error that manifests in event viewer is different.
Application java.exe has been blocked from accessing Graphics hardware.This is the error, which was previously the one that went along the lines of "igfxn has stopped responding and has successfully restarted".
There are further options along tweaking the registry that disable that TdrDelay completely, but I am too concerned about the risk involved. Letting the integrated GPU get itself stuck in a loop (potentially) sounds like a great way to nuke it, and I can't just replace the GPU in that instance.
Following further rabbitholes, all the advice given online follows the same as given here. Check the drivers etc.
I attempted to find some older drivers for my iGPU:
Intel Iris Plus Graphics (07 revision)
Intel Core i7-1065G7
BUTThe oldest drivers I could find (version: 26.20.100.7985) seem to date from march 2020 (so, this year!). I am an amateur at all of this stuff, so I don't know where else I can look, or even how to ascertain if there ARE older drivers available. And then if there are, how trust worthy are the sources. I am loathe to install something with that deep of permissions without being sure of the source.
So, once again, pretty much at square 1. Yippee! But I think logging some of the temps at crash may provide some insight, even though that may not be fruitful in creating a solution.
I hope some of my "research" may prove useful to the
powers that be, or to anyone who has a better understanding of computers. It's frustrating looking through event viewer to see things happening, but not really know what that actually means. Like when it says "java.exe has been blocked blah blah" the information that I can see doesn't give any useful info regarding what is doing the blocking, or how the blocking is happening or why the computer thought it necessary to block in the first place. Relying on inferences without a greater understanding is sure to land me in trouble, so I hope my scribbled notes of "what I have changed" will be enough to guide me backwards when I either run into massive problems due to tinkering, or land on a solution.
Cheers.