So I just took my last readings for AS5 vs MX-4 on the CPU-360...
My testing process was as follows using prime95 for loading and only showing 1 window on desktop to ensure I'm not adding inconsistent heat from GPUs
-15min @100% fan speed full load (ignore results)
-15min @100% fan speed idle (ignore results)
-15min @100% fan speed full load take measurement
-15min @100% fan speed idle take measurement
-15min @40% fan speed full load take measurement
-15min @40% fan speed idle take measurement
Temperature below = delta above ambient (or at least what I used to record ambients)
@100% Fan Speed
@40% Fan Speed
In short for what I did they are equivalent products at idle however under full load I was able to see upwards of 0.5'c difference in favor of good ol' capacitive AS5. However with two 6990's being in the same loop, and a radiator as big as the one I have there is always some margin of error. But running the tests multiple times I got pretty consistent results.
Ok so MX4 + CPU-380 and to be honest I ended up running prime95 for 30 minute cycles instead of 15 minutes because I simply didn't believe the temperature read outs.
100% Fan Speed
40% Fan Speed
I didn't break 30'c when the fans were on full and under full load.