FRONTPAGE G.Skill Ripjaws 4 DDR4-3000MHz 4X4GB Memory Kit Review

You really can't compare CPU bunchmark results across multiple CPU platforms when testing RAM.
It's horrible science and gives completely useless numbers. All you really see there is that newer, more efficient, CPU cores work faster, more cores are better, and clock speeds trump all in single threaded stuff.
It tells you squat about DDR4 vs DDR3.
Worse still, there's no information about the CPU clock speeds used. Given the extreme difference in CPU benchmarks with CPU clock speed changes, and the quite significant difference CPU speed makes in memory benchmarks, this is inexcusable.

I realize that there are no platforms that can run both types of RAM. That does not excuse horrendous science.
A one line disclaimer at the bottom of the results does not excuse it, either.

I'd hoped to see some DDR4 related information, maybe something about the edge connector redesign for better reliability and contact on flexible motherboards.
Instead all I see is marketing, shiny pictures and woefully deceptive graphs.

Thanks for the input, hopefully I'll have some more DDR4 samples soon and I can drop all comparisons to older platform stuff.

I wish to also see comparison at the same CPU clock and amount of cores in AIDA64 or similar multithreaded benchmarks like 5960X at 6 cores vs 4960X or 4930K. One simple reason is that memory bandwidth at higher amount of cores will be always higher on benchmarks optimized for multi cores. It's like comparing memory speed on Pentium vs i5 in AIDA64 or WinRAR.
The only thing that can be compared correctly are single threaded benchmarks like SuperPi.