Do you honestly expect 14nm++ to drop power usage by as much as 35% in the low 3Ghz range?
According to Intel, 14nm++ (Coffee Lake) will use 52% less power than the 14nm process (Skylake) or 26% more peformance, or any combination between. See page 4, top graph:
https://newsroom.intel.com/newsroom...es/11/2017/03/14-nm-technology-fact-sheet.pdf
Intel has already claimed that they were going for the sweet spot of about 15% more performance. Which according to that graph is 35% more power efficient.
So, a very rough estimate would be to take any 4-core Skylake, add 2 more cores (+ 50% power) and then use 14nm++ (- 35% power) and you are still right at about the same TDP (1.5*0.65 = 0.975 times the power from Skylake). That is a pretty close match to the results that have leaked so far:
- Skylake 6700K: 4 cores, base 4.0 GHz, turbo 4.0 GHz
- Coffee Lake 8700K: 6 cores, base 3.7 GHz, turbo ?
- Skylake 6700: 4 cores, base 3.4 GHz, turbo 3.7 GHz
- Coffee Lake 8700: 6 cores, base 3.2 GHz, turbo ?
- Skylake 6600K: 4 cores, base 3.5 GHz, turbo 3.6 GHz
- Coffee Lake 8600K: 6 cores, base 3.6 GHz, turbo ?
- Skylake 6400: 4 cores, base 2.7 GHz, turbo 3.1 GHz
- Coffee Lake 8400: 6 cores, base 2.8 GHz, turbo ?
As for multi-threaded performance, lets assume the 8400 turbo is the same 3.1 GHz turbo as the 6400 (I suspect that it may even be a tad higher). Assume a 95% efficiency in going from 4 cores to 6 cores. Thus (6 cores) * (3.1 GHz) * (0.95) = 17.7. Compare that to the multi-thread performance of the 4 core 7700K at 4.4 GHz: (4 cores) * (4.4 GHz) * (1.00) = 17.6. So, it wouldn't be surprising that the entry Coffee Lake i5 matches the 7700K in multi-threaded performance. The only difference would be hyperthreading with the 7700k which isn't that great of a help.
Edited for clarity (for pj- and IntelUser2000) since I skipped two lines and jumped to the conclusion.