Will Skylake-X be able to "rock our socks off"?

[TehPenguin]

I know it's release is still a year from now but bear with me for a moment, please.

Brodwell-E just jumped into the scene and the consumer range doesn't seem to have improved all too much over Haswell-E from 2014.

Skylake has joined the family last year and some benchmarks show that the i7-5775C outperformed it in many cases. Admittedly there's but a couple of weeks difference between Broadwell and Skylake but the point is: the newer i7-6700k does not offer much apart from slightly higher clocks.

Now I am by no means an expert and all I do is interpretation.

With that in mind: Skylake-X will be based on the Skylake die which AFAIK has an improved iGPU and more IPS over Broadwell. The real-life difference seems negligible to me.

Kaby Lake will join us Q4-16 but it doesn't seem to be improving much either.

 

[Sakkura]

Well, Broadwell-E doesn't have the L4 cache that was helping out the mainstream Broadwell chips in some tasks. So we should expect a little more of a gap per GHz (and per core) between the high-end CPUs than we saw between the mainstream CPUs. That said, Intel isn't making major performance leaps anymore, except when they add more cores. That pattern is almost certainly going to hold here as well.

Skylake-X is a new die, and shouldn't have an iGPU (just like Broadwell-E).

 

[TehPenguin]

OK, so going for the 6800k at the end of the year shouldn't make me regret it 6 months later. Thanks!

 

[Ecky]

Pretty much. Expect ~5-10% performance per generation at the same pricepoint. Skylake-X might overclock a bit better than Broadwell-E, which regressed when compared with Haswell, if DT Haswell vs Broadwell is any indication.