Memory timings, show me proof!

pragmatic

Distinguished Member
Joined
Nov 9, 2004
Messages
12,262
Reaction score
1,639
Points
1,968
I am sceptical about memory timings, I want to be proved wrong.
As it has often 2-3x more for the same amount of ram for the same supplier running at the same memory speed (i.e. 400mhz) but the only difference is the cas-las and other dodgy numbers for timings.

Now it is my personal belief that expensive lower timming memory (not dissing branded as that brings stability over generic and generaly wont brake as they usualy have a life time warranty) brings very little performance increase, say 2-3% percieveable difference in total performace.

I'm comparing this memory timings thing to speaker wire, and snake oil salesman, combined with general ignarance (unlike that this can be proven with impirical proof). I would like to be proving wrong and shown memory that lower timing when running stock will give me a 10-15% or more overall performance increase over lower timing "crap" memory, as this is what would happen if the equivalent was spend on a upgrade from midrange to high(ish) end graphics card, although ofcouse it would be more pronouched with a graphics card with 25-50% difference, in games of course.

This is at a stock speed, now everyone (should) knows that fsb is king when it comes to overclocking as the entire system is running faster not just a single component (and unless that component is the bottle neck the performace increase through oc a single compoonent would be minimal). So I wish to take overclocking out of the picture as a chip with better timings may have a higher max overclocking spead at which a system would be stable. If that was the purpose for buying the memory why not just buy some that is rated faster in the first place.

So is there or has there ever been a test between equaliy speeded ram (possibly form the same manufacturer) that shows that faster timings have a significant enough overall performance difference to justify there price. I.e. a cheap matched pair 400mhz 1gb (2x 512mb) may cost as little as £100 where as low timings 2-2-2-7 or some other crap may cost £200-250 or even £300 for the same speed and memory amount.

Anyway that is enough of the rant, i'm sure i wil be flamed but I want cold hard evidence. Also I understand If you have the highest end everything and need/want the 2% performance that the memory brings then thats fair enough i'm not having a go at you.

I would just like to say that I believe this is related to PC gaming (forum) as that is the use that this ram is for and i would like to show that the money could be wisely invested in a better gfx card, sound card (which do make a small difference and the price difference between them is not so much) and even a good hard drive over a slower one (although i don't believe in the raid thing in practice although thats another topic).
 
My advice would be to check out the memory tests on PC sites like www.tomshardware.com or www.anandtech.com as they usually comapre a number of chips at stock speeds as well as overclocked, in games and synthetic benchmarks
 

The latest video from AVForums

TV Buying Guide - Which TV Is Best For You?
Subscribe to our YouTube channel
Back
Top Bottom