ram speed vs cas latency

[Wishiwasdead]

Hypothetical question, I don't actually plan on buying any ram in the near future... but I was wondering as far as gaming goes what would be the best kit out of these three? 

 

Corsair Vengeance 

3000 cl 15

3200 cl 16

3600 cl 18

 

In the past (DDR3 days...) once you got up to a certain speed (about 1600 ish), speed didn't really matter anymore and a lower cas latency was more important... is that still true? Or did something change...

 

[Eigenvektor]

CL15 @ 3000 = 10 nanoseconds latency

CL16 @ 3200 = 10 nanoseconds latency

CL18 @ 3600 = 10 nanoseconds latency

 

So the absolute latency is the same, since that is a function of the CL number and the frequency.

 

[boggy77]

depends on the cpu as well. 

[boggy77]

1 minute ago, Eigenvektor said:

CL15 @ 3000 = 10 nanoseconds latency

CL16 @ 3200 = 10 nanoseconds latency

CL18 @ 3600 = 10 nanoseconds latency

 

So the absolute latency is the same, since that is a function of the CL number and the frequency.

 

latency is the same for all, in which case the higher frequency of the 3600 gives it an advantage. that's if the cpu can actually run the ram at 3600.

[Eigenvektor]

Just now, boggy77 said:

latency is the same for all, in which case the higher frequency of the 3600 gives it an advantage. that's if the cpu can actually run the ram at 3600.

Agreed. Latency being the same means frequency (=speed) wins.

 

To give a bit more info: The CAS Latency number is expressed in number of clock cycles. Higher frequency = clock cycle is faster = absolute latency is lower.

[trag1c]

15 minutes ago, Eigenvektor said:

Agreed. Latency being the same means frequency (=speed) wins.

 

To give a bit more info: The CAS Latency number is expressed in number of clock cycles. Higher frequency = clock cycle is faster = absolute latency is lower.

@OP For completeness I will point out that the real clock frequency of the ram is half of what the effective clock is. So 3600MHz ram is actually running at 1800MHz but because it is double data rate so you get 2 bits per clock cycle so double your effective clock.

 

So to calculate CAS latency you would use the real clock and not the effective clock.

[porina]

CAS latency is only one of many timings though, and the large array of others also have some influence in overall results.

 

Still, depends on the software being run, some scale better with overall latency, some scale better with overall bandwidth. I'm not up to speed on what modern games do when it comes to that behaviour. Differences are minor and it isn't going to make or break a system.

[Wishiwasdead]

Thanks guys, you answered my question!