RAM timing; does it make a big difference?

[MajesticHF]

I have 16GB of RAM (4 x 4GB).

 

Two of the sticks are DDR3-1333; the other two are DDR3-1600.

 

If I run all sticks at 1333MHz, I get 9-9-9-25.

If I run all sticks at 1600MHz, I get 11-11-11-30.

 

Does it make a difference at all? Would having the DDR3-1600 sticks running at 1333 MHz make it last longer?

[SpyRosL]

Nope.It's gonna be the same and it won't last longer if it's at lower/higher frequency. It's not a CPU

[H4X3R]

Best perf @ 1600mhz cl8 (for me at least).

I suggest you get whatever is cheaper and manually set the speeds to 1600mhz and CL9

[theduke33w]

at 1333 and 9 CAS you will have a cycle time of 13.5 ns.

1600 at 11 CAS = 13.75 ns.

 

you won't be able to tell a difference between those timings.

 

If you are using computer mostly for gaming I would suggest lower CAS Latency as gaming is takes more advantage of quicker Latency over bus speed.

[theduke33w]

You are going to want all four sticks to run at the same freqs. and latency or most likely the higher set won't be utilized properly.

[patrickjp93]

timings make a difference in applications so memory-demanding Intel's and AMD's caching algorithms (and cache space) can't keep pace. Generally, stick with 1333 CL 8 or better, or 1600 CL 9 or better. The clock rates bring greater bandwidth, and latency increases the downtime between a request for information and its departure for the CPU.

 

Mainly it matters in integrated graphics or in simulator programs where the CPU loads all the textures dynamically into the GPUs for processing. There are simulation engines and rendering scenarios where low-latency, high clock rate RAM will make a big difference, but chances are you're not using them.

[ManWithBeard1990]

The cheaper AMD chips benefit more because their cache is very small. In general, the lowest response time will be optimal, which is the CAS latency divided by the frequency. That's why 2133MHz CL9 will always be better than 1600 MHz CL7. For APUs bandwidth is important as well so you may want to optimise for frequency instead, but yeah, you're free to experiment and benchmark. For most cpus the difference isn't very big at all anyway. In your case, 1600 CL11 doesn't respond faster than 1333 CL9 (11/1600*1333=9.16) but the difference is so small that the increased bandwidth may make up for the small loss in latency. Honestly I doubt you'll see a difference.

[juretrn]

For gaming, none at all.