Is that fast ram actually doing anything for you?, say vs. 1600 @ 7-8-7-24?
Something about cycles per second. However latency takes it's toll.
So 2400 at CAS 11 would be 2400 cycles a second and would take 11 cycles to achieve.
OR 1600 cycles per second and would take only 7 cycles to achieve.
The 2400 is 800 cycles per second faster, but CAS is 4 cycle slower.
The 2400mhz would be only roughly 400 cycles per second faster because of the latency.
He could improve the latency to CAS 9 to increase the cycles per second and perhaps see a better gain over the 400 extra cycles per second.
I think I'm close on this. It's been a while since I've crunched cycles and shiznit.
But the technical answer would be the 2400mhz is faster by a small amount. Probably not noticeable in any real world application other than benchmarking. This would need some testing.
___________ A personal liking to speed____________
I generally try and run 1000 - 1100mhz at Cas 9. I've run many FX chips at 1100mhz at 9-10-9-27-36 1T and 2T depending on the sticks.
The BEST gains I've seen where from overclocking the stock 666mhz divider at 9-9-9-24-32 1T to 933Mhz OC.
___________A note with overclocking the CPU_________
Since the usual is to overclock the CPU up as far as it goes, this does have an effect on memory speeds. This generally holds true to video cards particularly NV. When you OC the memory, the CPU frequency at lower speeds yields better stability. This is all due to the memory divider at hand. When changing the reference clock or Bus of the cpu, it changes the memories divider and can cripple the overclock.
__________Opinion for overclocking CPU vs Memory__________
I'd rather have a much higher CPU frequency over the Memory frequency. It's the most important part in the rig.
Take that 4.6ghz CPU OC and 1600Mhz Cas 5-7 over 2400mhz Cas 11 Memory with a stock CPU any day.
