Temp questions...

[vudoo]

I have probes for the CPU, NB, and water. I have one for GPU also, but its just dangling outside the case while I wait for my vid card to come in the mail. According to Aquasuite, my temps are: CPU:24, NB:31, H20:24, and ambient (the GPU probe hanging outside the case): 21.5C. I'm not sure if I should be excited or worried. Speedfan is showing 41C for the CPU. I know speedfan isn't the best, but can it be off by THAT much? Is the NB usually hotter than the CPU? This is kinda making me think that I didn't place the temp probes properly. 3C above ambient for CPU sounds great though if it's correct...

 

[sadvocate]

I have probes for the CPU, NB, and water. I have one for GPU also, but its just dangling outside the case while I wait for my vid card to come in the mail. According to Aquasuite, my temps are: CPU:24, NB:31, H20:24, and ambient (the GPU probe hanging outside the case): 21.5C. I'm not sure if I should be excited or worried. Speedfan is showing 41C for the CPU. I know speedfan isn't the best, but can it be off by THAT much? Is the NB usually hotter than the CPU? This is kinda making me think that I didn't place the temp probes properly. 3C above ambient for CPU sounds great though if it's correct...

Location and calibration of the temperature probes plays a part in the difference you're seeing. If you have an IR temp sensor or another means of determining the ambient temperature, you can get a pretty good idea if the Aquero is calibrated correctly at stock settings by comparing the thermometer/IR sensor/whatever to the ambient probe (your unused GPU probe). You also have to consider where the Aquero temp sensors are mounted. Unless you remove the factory cap on the cpu, your Aquero probe is going to be a good distance away from the action - and unable to report the full heat exposure of the on-die temp sensor. Furthermore, there is going to be a small difference in temperature across your waterblock as well as across the stock thermal cap on the cpu.

Water temps are also significantly lower than the actual chipset temp. To illustrate, put an ice cube on metal so hot it is glowing (like the heating element on an electric stove). One side of the ice cube will melt quickly, but the other side will still be frozen. Depending on where you take the temperature of that event, you could get strikingly dissimilar results. So it comes down to what location is relevant, and in the case of chipset cooling, it's the temperature closest to the element that will fail (the chipset/heating element) when it exceeds a certain temperature.

Although complicated by the probe placement issue (perhaps you have a better probe placement on the NB than the CPU), you also have two different heatsinks, so the cooling efficiency of each heatsink has to be taken into account (especially if you're using waterblocks on both). If the NB heatsink isn't as good (probably the case) as the CPU heatsink, then it won't transfer heat as well and may have a higher temperature. On air, in the upside down mounting position that the Lian-Li V2000B case puts the mobo, my chipsets were between 5 to 10C hotter than my CPU, and causing stability problems thanks to it.

edit: Probe placement is the most likely culprit- as your cpu temp is the same temp as the water.