The monkey wrench is you said on a previous load it happened but then you found a fix. So I'm assuming by your wording your previous fix was some software driver related setting.
The symptoms by themselves are very much like hardware going south, there have been several product lines that have this symptom from the get go so faulty hardware design can be the case. The only way I can reconcile the two in my mind is the previous "fix" disabled a feature on the failing/buggy hardware so it wasn't an issue anymore. Check through the windows event log for any obvious hardware errors that might be a preview, though don't be completely surprised if you don't see anything as on a hard lockup Windows sometimes doesn't get the chance to write an event.
Try to cold boot into safe mode at some point (F8 after BIOS screen but before Windows logo and select safe mode) and see if it locks up in safe mode. If it doesn't then we'd lean towards the setting turning off some advanced feature was the previous fix. If it does then you've got a basic hardware issue.
As time and spare parts permit I'd do a process of elimination on the hardware side of things. Get every thing down to the bare minimum needed to boot windows and work from there. That's Power Supply, Processor, RAM, Video, Keyboard/Mouse and only the system drive then check there. If that freezes then one at a time swap parts in and out, if it doesn't look towards either one of the parts you unplugged being bad or the Power Supply not able to handle the full load from cold. If you have a spare power supply I'd start there in either scenario, if not go with the memory. On your memory you want to go one stick at a time through each slot. So stick 1, slot 0, then move it to slot 1 etc til all slot/stick combinations are tested. You might find one of your sticks is bad or one of your slots is bad or both or some flaky combination where slot 0 really hates stick 2 etc. Work through your hardware this way as best you can to see if you can find the culprit. You can also disable on board hardware one at a time in the BIOS as tests to those.
If on the Safe Mode test things don't lock up and you just want to find the fix you had before. I'd turn every sleep, hibernate, and power save setting to off. Go into things like your video card settings and make sure no overclock features are enabled and start one at a time disabling any advanced features. I'd also disable/uninstall any "helpful" tuning/monitoring apps.
Assuming you did this but a fast way to see if it's related to your specific hardware being a buggy design is to google the model numbers and cold boot freeze etc. You might find some specific answers like some of the first G skil triple buffer ram, certain Asus boards etc that have had these issues in the past.