Is this even possible?

[tolis626]

Hey guys, I've been messing around with my system today, and I changed some settings in Asus' AISuite software, which usually works quite well.

So I bumped up my uncore clock and voltage, switched to manual voltage for the CPU and headed to do some stress testing, until I saw this gem.

So I went to the GPU tab of AISuite and saw both sliders set to max. I hadn't touched my GPU settings (I'm stuck with the iGPU for now, so I've overclocked it to 1600MHz with 1.2V to casually play some Borderlands 2) for a while, but they were always at what I had set it to in the BIOS. I've seen it happen again, but it hadn't been applied. So, as you can imagine, I almost crapped myself. I haven't even rebooted yet, as I was in panic mode for a few minutes. It seems to work though, now that I have toned it down to my usual settings.

I'll do a reboot and see what happens...

My question is, is it even possible that this really happened? What bugs me is that HWMonitor measured it, not AISuite. I just suppose that if anything got run at 3V, it would catch fire, not just stop working. :'(

 

[tolis626]

Aparently, the BIOS can "only" supply the iGPU (and core and uncore for that matter) with up to 1.92V. So I think that 3V was just a false reading and never got actually applied on a hardware level. Which seems logical as at 3V I'd expect flames to start coming from the chip. It also works well, so no damage done. I don't feel like testing again though. Having a heart attack at 21 is bad.

 

[monkeydelmagico]

I've found HWMonitor to be off for some of my 4690k readings on an Asus z87 board. Still learning the AISuite myself, prefer to do most of the tweaking in the bios.

 

[tolis626]

I've found HWMonitor to be off for some of my 4690k readings on an Asus z87 board. Still learning the AISuite myself, prefer to do most of the tweaking in the bios.

I can agree with that. Another example is IA voltage (I don't know what the hell that is, though). When I use a positive offset with adaptive voltage, it will stay at something like 0.02V above the max VID. If I switch to a negative offset (We're talking about +0.001V and -0.001V here), it will skyrocket to something like 3.384V. And that's 100% reproducible. Strange. For now, I'm using a positive offset and am happy. For some reason, I can't have my offset set to 0. Whatever...

 

[monkeydelmagico]

forum member deders turned me on to Argus Monitor. Free trial and works well. Doesn't seem to suffer some of the glitches that HWMonitor does.