Worst case scenario for 45W TDP Skylake-H, and already faster than 65W TDP Broadwell-C/K. At equal TDP it should perform better as well. Let's see if GameGPU's garbage with old/unspecified Intel drivers and crap memory reflects what reviews will show this week.Agree.
My results are probably worst-case. Ambient temp was rather high (23C), the NUC is mounted directly on the back of my monitor, close to the wall, and I ran the tests all in a row. I like stress testing, so I would rather know the lower scores vs. just the best case. Will try doing some individual runs later for best-case results.