0
$\begingroup$

I was looking at the Wikipedia page for CAS Latency: https://en.wikipedia.org/wiki/CAS_latency

And I noticed the obvious trend - at higher frequency, despite CAS latency degrading (growing), the delay to get the 1st, 4th and 8th word overall improved regardless.

I've heard in the past that frequency/CAS was perhaps the best metric for memory performance as a consumer, but this chart seems to disagree:

enter image description here enter image description here

So my main question(s) are:

  • Is getting the 1st, 4th and 8th word an accurate representation of real-world performance of memory? If not, why not, and what would be?
  • If there some other factor to memory performance that would be a third "hidden variable" that I'm not accounting for, that's not constant between all these? The IMC's performance at high frequency? The motherboard's? Signal noise?

Thanks for any thoughts!

$\endgroup$
0
$\begingroup$

Another complicating factor is the use of burst transfers. A modern microprocessor might have a cache line size of 64 bytes, requiring eight transfers from a 64-bit-wide (eight bytes) memory to fill. The CAS latency can only accurately measure the time to transfer the first word of memory; the time to transfer all eight words depends on the data transfer rate as well. Fortunately, the processor typically does not need to wait for all eight words; the burst is usually sent in critical word first order, and the first critical word can be used by the microprocessor immediately.

Guess where I found this explanation

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.