Originally Posted by Joshua89
As I understand, the processing of information ultimately takes the same amount of time between "7-8-7-18" and "6-8-6-20." My system is stable at both, but no "tighter". "Tighter" in the sense that both sets of numbers add to 40. To put it in an anology, if two people are 6ft tall and are trading length of torso for length of legs, is there really any benefit one has over the other in the end if both people are 6ft tall in the end anyway?
yes and no tbh, they are all closely linked to oneanother, and although the summation of 7-8-7-18 and 6-8-6-20 are the same, it does not correspond to identical memory performance in all areas.
For example, your IMC wants to access X3 (row X column 3), so it waits for the row delay (tRAS) before it looks for X, then waits for the tRCD (RAS to CAS delay) then, if CAS is the same value as tRCD, which it usually isn't (about 2 higher) it looks for column 3 and executes a read.
The thing is, changing tRAS (which should be CAS+tRCD+2ish..higher for higher frequencies) makes very little difference in performance when compared to chaning CAS or command rates.Edited by fluxlite - 5/1/11 at 2:34am