For our tests we used an open air testbed with the following configuration:
We tried to produce the the highest possible temperature inside our test PC using the OCCT v2.0.1 software with a custom 30mins (mixed) operation.
We left both CPU cores running at full load for ~30 minutes. All temperatures were logged using the CoreTemp 0.99.3 and the software. We also followed two overclocking scenarios:
We measured the produced noise with a Precision Gold N09AQ Environment Meter. All measurements were made at a distance of 30cm away from the cpu cooler running at full speed. For all the coolers we used the same thermal compound from Tuniq, the TX-2.