Dear all:
I am on the case.
We knew about the problem with "W" and CPUs for a while, right? If you look for posts by groth he is working on an ammeter. With dumping that (current) and voltage into some of the inputs of my Maxim setup, you could in theory know C/W by simultaneously measuring A,V, and deltaT with a reasonable accuracy. In practice there are still some issues with the circuit, and the motherboard modifications are not for the faint of heart.
I have a die simulator and a heater cartridge and just need to drill a couple of holes in it. It is identical to Joe Citrella's at overclockers.
JoeC, BillA, and I have discussed cross testing our simulators (same block on all systems). We can then adjust using Bill's setup as the reference. In theory (damn there it is again) we can cross correlate somewhere down the road.
Basically this is way out there and more for sites not end users. Hey you don't go crash your car to test out the airbags and safety features do you? You let people whose job it is do that. My job here at procooling is to test waterblocks and rank them comparatively (and I am doing a shitty job of it so far).
For end users, we JUST posted a thread about putting a thermocouple under core. Thermistors work too and are much cheaper. Make sure to calibrate whatever you use before you epoxy it down. Good enough for you to tell whether cooling changes are for the better or for the worse. As far as comparing with one another, well that's just a big dick contest to start with; post wCPUID screens or pictures of blue/UV lighting action instead.
