Timings and La­tency


THE NEXT part of the holy trin­ity of mem­ory spec­i­fi­ca­tions re­volves around timings and la­tency. There’s a ton of them, but the most im­por­tant one you need to keep in mind is the CAS la­tency. Re­fer­ring to the Col­umn Ad­dress Strobe, this fig­ure in­di­cates just how many clock cy­cles it’s go­ing to take for the mem­ory mod­ule to ac­cess a par­tic­u­lar mem­ory lo­ca­tion, ei­ther to store or re­trieve a bit of data held there, ready for pro­cess­ing by the CPU.

That said, this fig­ure on its own doesn’t give you all the in­for­ma­tion you need. It’s only when you com­bine it with the mem­ory trans­fer rate we men­tioned above that you get a bet­ter pic­ture of just how fast your mem­ory mod­ules are. So, how do we get a fig­ure that makes any sort of log­i­cal sense to us con­sumers? Well, there’s a handy for­mula that con­verts CAS la­tency and MT/s into a real-world la­tency: La­tency = (2,000/ Y) x Z, where Y is your RAM’s speed in MT/s, and Z is your CAS la­tency.

So, as an ex­am­ple, if we take a 2,666MT/s mem­ory kit, op­er­at­ing with a CAS la­tency of 15, we get a real-world re­sult of 11.25ns. This tells us ex­actly the to­tal time it takes for that mem­ory mod­ule to ac­cess, store, or re­quest a bit of data from a lo­ca­tion on the mod­ule.

This is where over­clock­ing mem­ory typ­i­cally comes un­stuck. Gen­er­ally, the higher the frequency, the higher the CAS la­tency, and as such, real-world per­for­mance gains are of­ten slim, un­less the rise of that CAS la­tency is slowed as well. When it comes to the “best” per­form­ing mem­ory, what you’re look­ing for is a kit that has a high frequency, a low CAS la­tency, and the nec­es­sary ca­pac­ity (see op­po­site) to do what you want to do.

Newspapers in English

Newspapers from Australia

© PressReader. All rights reserved.