Overclock.net banner
21 - 23 of 23 Posts

· Registered
Joined
·
3,808 Posts
Discussion Starter · #21 · (Edited)
If your temps are that high them you should definitely redo it, and maybe use a little more LM. Post pics of what it looks like when you open ur CPU back up.

At those temps your CPU should be throttling so you are probably already losing performance.

I have found the best thing to do with delidding is use some clear nail polish to cover the PCB on the CPU that is close to the die, that will protect the PCB and the nail polish won't hurt anything. Then you can feel more secure that you won't mess up your CPU being more generous with the LM. Also, what LM are you using?
I use a cleaning solution each time i delid, Articsilver branded stuff i think?? A "Thermal material remover" (GREAT for removing the old silicone glue for the IHS) and a surface purifier. (I also use rubber gloves). Regarding thermal throttiling - it actually doesn't. I monitor it via "Argus monitor" (Love this program) and i get absolutely no throttling at all, but the high temps cause serious instabilities and/or even crashes. I Don't think iv'e seen it throttle until it's exceeded at least 105c


I've already done the delid using a different technique; i applied the liquid metal to both the copper IHS and the core (Being generous on the core side) and temps have decreased even greater then past applications.


However, after doing this, and booting it all up, i've ran into a bigger problem; my 1070 zotac GPU is no longer recognized. I removed the motherboard entirely and reseated every connector possible (GPU ((duh)), RAM, CPU power leads, mobo leads) and still the motherboard doesn't detect the GPU. Currently running off integrated at moment.

I'm thinking either the card has suddenly up and died, or the PCI/Mobo is fubar'ed. Either way, it's gunna be a costly fix - not having much luck am i?

I've ordered a very cheap PCI-E gpu to run a a test to determine what has failed. I'm going to hazard a guess it's mobo as thats the only thing possibly disturbed in what i've been doing. dont understand it either as i'm rediculously thorough and take my time with any work on my main rig. Bloody typical huh?
 

· Graphics Junkie
Joined
·
4,520 Posts
Check for bent pins or minor debris in the CPU socket. If your gpu is not being recognized after taking the CPU out and putting it back in, you may have bent or blocked one of the pins in the socket that carries the PCIe lane signal.
 

· Not a linux lobbyist
Joined
·
2,938 Posts
Also try with your old, stock ihs again. I also have one of those unplated copper ihs's from rockit and I shaved the bottom down a bit since I wanted to make sure the copper contacted my silicon well. It could be a touch thicker like mine was and elastically bend a pin or two a bit too far so it doesn't contact.
I also don't use it since, while it did help temps a bit, I was worried that I would have more lm related maintenance with it. Mine seemed to soak up lm even after I intentionally stained it. But it is beginning to sound like I could also reduce my maintenance by switching from coolabs pro to coolabs ultra as well.
 
21 - 23 of 23 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top