Maybe I'm just too old school, but back in the day when I was getting game crashed due to overclocking or some other BIOS setting, I would run stability test software and tweak settings to get to a point of high confidence in system stability.
What I see now is people running out of the box motherboard overclock settings (MCE XMP and the like) and being completely oblivious to stability problems that exited right away.
Yes, I'm aware of the degradation problems and how that plays a big role in all this. I'm just saying that people had these CPUs pushed beyond their limits from the get go and degradation worsened the problems they already had over time.
Now we have games with code (shader compilation) that sort of acts like a stability test putting the issues front and center right in their faces.
There are also those that were on the very edge of stability with the "hot" out of the box motherboard settings. The slightest amount of degradation would make the system unstable.
What I'm seeing people do is try workarounds for the crashing rather than apply BIOS fixes/settings. It's ignorance building upon ignorance.
It's clear that the CPUs were pushed too far by both Intel and the motherboard venders who wanted to see "numbers go up" in reviews
While I don't disagree with the gist of your statement, as someone who has built many (relatively for just a guy) PCs its been near impossible to know what it going on for some time now.
I think the first time Igor responded to one of my posts I was talking about building with Intel on a 10th gen but enabling XMP did who knows what in addition to setting the timings and speed of the ram to the hot profile (IMO, this should be all it does. ram voltage, speed and timings based on the stick data) so I would manually set the P1 and P2 wattages and durations because most of the time, out of the box, the Intel boards had P2 set to the moon and for an unlimited or extremely long duration. And then I'd memtest 24 hrs then Prime for 24 hours then loop Unigine Heaven overnight then stamp it good if it cleared all the hurdles. That's really similar to the burn in Falcon NW did to my first PC back in ~2002 and that seemed like a reasonable test to me.
And Igor was like, why set those power limits? And to what?
For the settings, great question, I had to reference ARK every time I was building an Intel PC because the bios settings never seemed to be what Intel specs were to get this data.
Ridiculous.
And for the why? For me it was the best and quickest way I knew to keep these boards from doing some runaway craziness (in all fairness, the voltages could still go ape and I guess this wouldn't help much) on these CPUs since the firmware had moved so far away from "stock settings" by default. Also, I like quiet builds with nice fans and keeping the boost power reasonable is all part of controlling peak fan speeds & noise.
And here we are. I am not surprised in the slightest that many people building their own PCs enabled XMP, set the boot device, saved and exited the firmware interface never to visit it again. That should be safe. That should be reliable. In no way should we blame the casual PC builder for not being able to trust the configuration out of the box. This should have been part of Intel's firmware guidelines since we moved to UEFI (if not earlier) and I'll die on that hill.
Please don't kill me.
www.hardwareluxx.de
Facebook
Twitter