I don't know what's happening to my GPU, last week it worked fine. Today, while running a Borderlands session, i suddently want to overclock a bit. I used hwinfo to log the temp before proceeding and found out my MemIO jumped from 62 to 128 (C degree) and then cooled to 70-ish. I decided to repaste and it reduced to 110 at max. It's bad that i don't have any thermal pad with me right now. I think changing the pads is a good idea, but could some going-to-be 6 months old pads cause this??? Or could hwinfo be the one at fault here??? B/c as i remember jumping to 100+ will cause the comp to shut down, but in my case, the game ran smoothly as normal
Update: it seems that hwinfo is not at fault here, GPU-Z says the same thing about the temp. I'm off to order thermal pads then.
-
ElectricTurtle Notebook Consultant
-
confirm it by running a few other temp monitoring programs, but new pads are always recommended.
-
The MemIO is the memory controller, which is on the die. The VRAM modules on mobile cards don't run nearly as hot as most people think.
Make sure you have your card pasted properly, it's possible that the memIO is located at the edge of the die, which sometimes doesn't get covered properly. -
ElectricTurtle Notebook Consultant
Update: i repasted and found out that the plate and the thermal pads below it are not spanning all of the VRAMs. I stretched the pads to cover them and the result shows max temp is 81, still high imo, but it's safer now. I'll have to replace the pads :-< then. Thank you for all of your help guys. -
Joe -
@ElectricTurtle what gpu do you have?
My 5870m cards peak on 88c MemIO (gpu0) 86c MemIO (gpu1), so 81c is perfectly fine if it is an ati card.
If you have a nVidia card 81c is a bit high though. -
ElectricTurtle Notebook Consultant
Well. The main reason why i said it's still high bc i used to have better temps in hotter room degree (it was about 71 if my memory serves me right)
MemIO jumping to 128 degree @_@
Discussion in 'Alienware M15x' started by ElectricTurtle, Jul 5, 2011.