Reported.
Anyone can measure Surface 3's power consumption, it's not something forbidden like the Mullins reference platform AMD sent to reviewers to make it look competitive inside a 11.6'' thick chassis that doesn't resemble any typical 7-8'' tablet Bay Trail-T product. They obviously had to hide the outrageous power numbers and put a giant chassis to allow it to sustain Turbo speeds even though they knew it would fail miserably to deliver this kind of performance inside a real small tablet, you know, similar to your Core M accusations. OEMs obviously knew that, hence why it almost didn't make it to the market. Do you really think they couldn't have launched such a product at higher prices if it really outclassed BT like that? Surface 3 at $500 is the proof they could.
Ps: Braswell completely smokes your precious Kabini in perf/watt, even AMD-biased Hardware.fr aknowledges that:
http://www.hardware.fr/articles/imprimer/936/
LOL...
You realize that you are denying that Contra revenues are the cause of Mullins being not present in W8 tablets..?..
I stand by my post that whoever say otherwise is intellectualy dishonnest.
As for Hardawre.fr review i did read it, cherry trail smoke nothing, CPU perf/Watt is not improved, quite the contrary, GPU is much improved but TDP is understated by Intel as proved by this review that yield roughly 10W at the SoC level for the higher grade part...
All they did was to reduce frequency by 10-20% because this improve mechanicaly the perf/Watt by the same ratio, that s completely irrelevant as comparison wich must be done at equal frequency and electrical environment...
Edit : As pointed post the infos in your btrail thread, why keeping derailing this one..??.
You are afraid that i wouldnt answer there..?.