I have an update on the subject of Nanofluid. I got a new sample (new formula) of the fluid for testing.
I decided to do a very down-to-earth practical test of the fluid this time. I used it in my business PC, equipped for this purpose with a GTX 285, so that it produces a bit more heat.

Here are the components making up the loop:
- XSPC X2O 450 pump
- Watercool 360 radiator
- D-Tek Fuzion V2 (on a Q6600 @ 2.7 GHz, 1.3 V)
- EK FC-285 GTX (on a PNY GTX 285 @ stock)
- 1/2" tubing
- Some thermal probes

As you can see, the loop has a relatively underpowered pump (nice and quiet, though =) ) and a radiator that is actually a bit too small for the CPU and such a powerful GPU.

For the test, I first ran the loop with distilled water. I ran Prime95 (small FFT) and Furmark for 45 Minutes and measured the temperatures. Then, I changed to Nanofluid in the loop, did the same test and measured temperatures again. Here are the results, shown in temperature-differences between component and fluid temperatures.



As you can see, the Nanofluid does make a measurable difference, here. I must add that the error margin for this test is expected to be higher than for the tests I run on my test-bench, since I don't have the same, precise measuring tools in my working PC. The results still show a positive trend, though.

I am now running my PC with the Nanofluid and intend to keep it like this for at least a few months. Then, I'll take the loop apart to see if there are any long-term impacts from using the Nanofluid.