Temperature difference between cooler and cpu

Tcamp · May 23, 2015

Hey everyone,

I am wondering what temperature differences should you see between your CPU and water cooling unit. I have a 4690k CPU overclocked to 4.5GHz using Asus AI, an H100i cooler using the pre-applied thermal paste from factory, and I am using Corsair Link to monitor temperatures. Also, this is in a room about 22°C. To explain why I'm wondering what temp differences you should see; my cpu temp drastically rises when in heavy use but my cooler doesn't seem to rise much at all. At idle my temps are fine as I am typically around 30°C and my H100i temp sitting at about 28°c. I do sometimes notice the cpu temp jump around a little more now (2 weeks after installing the cpu and Corsair Link) but no higher than 5°C jump while opening a new program or something like that; and to add, overall my cpu temps are not bad because I have never seen it rise above 52°C and only that high when using EVGA's cpu burner to acheave 100% load. What makes me wonder is when I do go to place a load on my cpu I see it's temp rise quickly, say up to 45°c more or less, but my H100i temp doesn't seem to rise along with it all that quickly. In fact I've never seen my h100i temp go above 35°c and that was only when using EVGA's cpu burner for a few minutes and my cpu resting about 50°c for that time period.

I'm cutting myself off here as this is already too long but hopefully you get the point of my question and understand if there is anything I didn't specifically ask but you feel is important then do mention.

Thanks everyone,

Tyler

Speaker1264 · May 23, 2015

Try running your cpu at full load for over an hour and then see. It is going to take a long time for a cpu to heat all the liquid inside your cooler, and the cooler is made to dissipate that heat very quickly, so this is not unexpected.

KemoKa · May 23, 2015

It's just thermodynamics. Water doesn't heat up instantly, and that CPU is actually dumping a lot more heat into the cooler than the difference in temps is letting on. The difference in temps is just the difference between how quickly your CPU is producing heat and how much heat the rad is dissipating. leave it for a while and the cooler will probably catch up

Tcamp · May 24, 2015

It's just thermodynamics. Water doesn't heat up instantly, and that CPU is actually dumping a lot more heat into the cooler than the difference in temps is letting on. The difference in temps is just the difference between how quickly your CPU is producing heat and how much heat the rad is dissipating. leave it for a while and the cooler will probably catch up

Try running your cpu at full load for over an hour and then see. It is going to take a long time for a cpu to heat all the liquid inside your cooler, and the cooler is made to dissipate that heat very quickly, so this is not unexpected.

Thank you. However, is the temperature difference between my CPU and cooler (when it is under load) normal? I know the water will absorb a lot of energy do the hydrogen bonds so I would hope my cooler doesn't rise to 50°C that quick but I just figured my water block would have be more thermally conductive, than letting my cpu get up to 50°C pretty much instantly when using the cpu burner. If that all makes sense... I considered the thermal paste not performing well but I don't know what to look for when specifically dealing with thermal paste.

Speaker1264 · May 24, 2015

Thank you. However, is the temperature difference between my CPU and cooler (when it is under load) normal? I know the water will absorb a lot of energy do the hydrogen bonds so I would hope my cooler doesn't rise to 50°C that quick but I just figured my water block would have be more thermally conductive, than letting my cpu get up to 50°C pretty much instantly when using the cpu burner. If that all makes sense... I considered the thermal paste not performing well but I don't know what to look for when specifically dealing with thermal paste.

The cpu block measures the temperature at the far end of the block, so it isn't not measuring the direct contact of the block to the cpu, but basically the temp of the water that is passing through. CPU's work exactly how you describe, they get rise and drop very quickly because the sensors are extremely small. If you were to run a stress program that utilizes 100% of the cpu for over an hour you will see the cpu temp and the water temp get much closer, but they will never be the same because the radiator is dissipating that heat.

KemoKa · May 24, 2015

Thank you. However, is the temperature difference between my CPU and cooler (when it is under load) normal? I know the water will absorb a lot of energy do the hydrogen bonds so I would hope my cooler doesn't rise to 50°C that quick but I just figured my water block would have be more thermally conductive, than letting my cpu get up to 50°C pretty much instantly when using the cpu burner. If that all makes sense... I considered the thermal paste not performing well but I don't know what to look for when specifically dealing with thermal paste.

The temps you're getting are totally normal, especially given how fast the CPU is going. A lot of people nowadays have started scoffing at 4.5GHz for some reason. I have my 2600K at "only" 4.0GHz because of impending summer but it's still crazyfast. 4.5 is a mad overclock, and the temps you're getting are actually really really impressive. CPUs only start getting in trouble once you get to about 90°C, 50 is really nice.