SOLVED: GTX 1070 running hot with high clock speeds.

takochako

Prominent
Jul 25, 2017
50
0
660
I was streaming TV on my computer today and I noticed that there were some artifacts happening on my other monitors. The image on my monitors would stutter and slow down then all of a sudden speed up to catch up with the current frame. I decided to check PrecisionXOC and MSI Afterburner to look at the CPU usage and noticed that my graphics card is running steady and 987 MHz GPU clock and 4006 MHz memory clock. The temperature is very stable and is either 43 or 44 degrees Celsius. GPU usage is around 0%, with some minor spikes up to 1 or 2%. These are all idle stats and I really don't like the fact that it's doing this. All of this carried over restarts and power cycles and I have no idea what else to try. I read somewhere that this could be caused by spyware, but honestly I doubt it. I have Malwarebytes pro and the latest version of ESET Smart Security running at all times. I also have a Pi-hole set up blocking a whole bunch of malware, ransomware, and advertising domains (yes, I am a bit insane when it comes to security on the internet). Does anyone know why this is happening and how to stop this?

UPDATE 2/7/18: This has been solved, could a moderator please mark the answer I made with the link to the EVGA forum as the answer?
 
Graphics cards can get very hot. My old Titan X would feel like touching a hot kettle if you put your hand on it while gaming. When I got my 1070, I bought a 3rd party gpu bracket for it and installed a water cooler. Before the water cooler, the 1070 would also get really hot to the touch. With the water cooler, that hasn't happened until the new Assassin's Creed Origins game. If I play on 4k w/ ultra settings, the card now gets pretty warm(although the game does play smooth).
 


Until yesterday, this never happened. The card usually ran around 300MHz at idle, and the idle temps were between 20C-30C. It has never gotten above 60C under a heavy load.
 
Have you aired out the system recently? It would also help to check the thermal paste on the GPU and make sure it isn't dried out/cracking. Although if you are going to inspect, you may as well replace the thermal paste with something quality. I prefer using Grizzly Kryonaut(I use it on all my CPU's & GPU's, desktop and laptop)
 


By aired out do you mean just opened up? I do that pretty regularly since the side panel on my PC is pretty much a door. I'll check the thermal paste, but that still doesn't explain the high idle clock speeds.
 


I spayed the entire card with compressed air just 2 weeks ago. I'm more concerned with the fact that the card is running at 987MHz or 961MHz at idle and not fluctuating apart from when I open snipping tool or GeForce Experience. It jumps up to 1594 when I open any program, and then jumps back down to 987 or 961. The memory clock doesn't change at all. When I opened PrecisionXOC, it seemed to have remembered that I had my fan curve set to aggressive, and the fans sped back up which brought the temp back down to around 45C after about 5 minutes. Here are some screenshots from PrecisionXOC. I don't know if they'll help at all, but I might as well include them in this post.
Screenshot 1: opening snipping tool
Screenshot 2: Idle
 
I changed the thermal paste and the card is definitely staying cooler now, sitting around 36C. The thermal paste was NASTY, it had turned into a soft solid sort of thing. The GPU and memory clocks are both sitting right where they were before though. I've noticed that every once in a while the numbers showing the clock speeds blink out for about a frame or two, and the blink back in, the same as they were before. I'm pretty darn sure this is a software issue, but I don't know which piece of software could be causing this. I've been running Malwarebytes for about 14 hours now (I'm scanning 3 hard drives, including a backup drive) and it hasn't found anything other than Claymore's Miner (don't judge me, I was testing it out xD) and OpenCandy adware from an emergency backup sorta thing I made of my old laptop. The fact that it's running with such high clock speeds at idle makes me wonder if I have some sort of cryptocurrency miner that was dropped by something. But then again, I have been reading a lot about miner trojans lately and I get kinda paranoid about these things, so while it's possible, that probably isn't what's happening. Does anyone have any idea of what would be causing the high clock speeds?

UPDATE: I think I found the cause! I have two Samsung C22F390 monitors plugged in via DP, my Oculus Rift, and one Dell S2330MX monitor plugged in, also via DP. I was experimenting around with different combinations of monitors and found that when I unplugged the Dell monitor, the clock speed jumps back down to 150MHz at idle, and the memory clock speed drops way down to around 400MHz at idle. The temperature quickly dropped by about 5C. Does anyone know why this monitor would make my GPU do this but not the others? I've been using it for about a year with no problems until recently.
 

TRENDING THREADS