I have a i7-8700k cpu and a 1080ti gpu. However when I run any game with even moderate graphics it seems that the CPU is doing all the work, heats up and my liquid cooling fan and pump go into overdrive and make a ton of noise. Even running something like Conan Exiles, after a few minutes my cpu cooling system is out of control trying to keep the cpu cool. Shouldn't the 1080ti be enough to handle the burden of the graphics without using so much cpu power?
My drivers are all up to date. Maybe a setting somewhere?
Comments
Assuming that the monitor is plugged in to correct port, then check problem game's settings. Try to find setting called "Frame Rate Limiter", and set it to enabled. Games may also call this setting Max FPS, FPS cap, Frame Rate Cap, or something similar. If your frame rate is not limited, then your computer is calculating as many frames as possibly (even beyond your monitor's ability to display them), and working as hard as it can. This is not an error, it's intended design. But in many games you can keep the volume down by enabling the Frame Rate Limiter and thus having the game limit the amount of work done by your GPU and CPU instead of using maximum resources available.
If the liquid cooling is making a lot of noise, you might also want to check if it's installed correctly, working correctly, and the settings are correct. If it's not working properly, or if it's trying to keep your CPU cooler than necessary, something like that could create a lot of extra noise needlessly.
You would definitely notice if the CPU were doing the graphics - it would run like a slide show, with the absolute minimum of detail on really low settings. It's an either/or thing - either the CPU is doing the graphics internally, or the 1080Ti is doing the graphics. They won't share or only do part way.
Where did you buy the computer at?
What does Task Manager and/or Resource Monitor say when you are trying to do something - it will show % loads for CPU/GPU/etc. You would need a different app to show temps, but that would be a really good idea in your situation.
I built the computer and upgraded over the years, wasn't a store bought pc. It does the same thing if I use my htc vive headset. Everything looks and plays just fine, it's just the noise from the fans ramping up so high that bothers me.
The frame rate limiter sounds like it could be the reason, but if it's application specific and I have the same issue on multiple games, I doubt every game I've played has that option. Maybe a global max fps setting in the NVidia settings (if there is one), or a 3rd party program could resolve it.
I've done that as well. I'm currently running BDO and the cpu usage is 19%, gpu 22%, and the fan has sped up to 2700rpm and the pump 1800 and I can hear it pretty good now, not as loud as it gets sometimes, a few more minutes and I'm sure it would get louder. I'll try checking my cpu paste and replacing it again.
That sure sounds like something about your CPU cooling system is done wrong. It could be not physically connected properly, causing it to run too hot. It could be improperly configured so that it turns the fan way up when it shouldn't. It could be something misreporting the temperature entirely. But 43 C at idle is way, way too hot for a modern CPU unless the ambient temperature is really high.
I'm not sure what to try next.
Take the fans off and blow the radiator out - may even run that part under water to get it clean.
Strip the old thermal paste off of everything - the water block and the CPU. Everything needs to be pristine clean. New thermal paste - it goes on as thin as you can get it and even, like melted butter on toast -- not like peanut butter on PB&J.
Put it all back toghether. If your idle temp is still over 35C (and do make sure your idle - Task Manager should be showing <5% utilization and very likely should also be at a very low clock speed), and the pump claims it’s running and the fans are spinning - then your 100i is bad and needs to be replaced. Intel chips will idle <35C on stock HSFs, and should be running much closer to 30C on a AIO water cooler.
거북이는 목을 내밀 때 안 움직입니다
That's what I did. Cleaned everything and put new arctic mx-4 thermal paste on. The H100i v2, is about 3 months old, as is the cpu and mb. Right now the load is 1% and the fan and pump are running around 1800-2000rpm with a temp of 35.1c. The cpu however has a temp of 43c.
I can use the Corsair link software and speed up or slow down the fan and pump, change the led color, etc. So it seems to be hooked up right as far as the controls go.
That thing running at idle, the fans shouldn't even have to spin at all, the pump runs at more or less a constant speed, and you should be under 35C on the chip.
The Corsair will be reporting fluid temps, the chip is obviously reporting chip temps - so they should be slightly different from each other. At idle they should be pretty close together (within a few degrees), but they will diverge a good bit under load. That applies for a good air cooler as well. If you have a really aggressive overclock with overvolt, and have the power plan set to Max Performance - maybe, but even then I don't think at idle you'd see this much temperature deviation if the HSF and everything are working correctly.
Your temps sound normal for a CPU under low (~30%) load... but at idle everything should be sitting at just over ambient (30-35C is normal inside a computer case for typical room temps - ~75F). EIther your CPU is still crapped up with malware and it's not really idling, or you haven't installed the H100 correctly, or your H100 just isn't working correctly. I'll take you at your word that your system is now free of malware and you installed it correctly. Not really rocket science and there aren't really a lot of other options there that I see that would affect a system this much while idling.
Nope, no OC at all. I have about 8 fans in the system, some blowing into the case, others out. It's a full size case with a lot of room. I'm going to try changing the radiator fans since they are noisy as hell anyway.
A couple of degrees here or there isn't unheard of, but nearly 10C is a bit excessive for that. And there are other thermcouples on the dies - you can see individual core temps with some software, they should all be in a pretty tight grouping depending on load.
Has the chip ever run at normal temps, or has it always got hot?