• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

RAM Timings and MHz

alimoalem

Diamond Member
the MHz is pretty self-explanatory but adding in the Timings is what gets me. what do the timings mean? why is 2-3-2-8 better than 3-3-2-8? and if higher MHz ram generally has worse timings, how is it faster? any help would be appreciated and thanks in advance
 
Holy sh!t, I used google and it was the first entry:
http://www.andrew.cmu.edu/user/fma/P4tweakRAM2.htm

Even though it has to do with OCing ram on a P4, it gives an explination of everything. I got another link somewere about ram timins, I'll see if I can't find it. And you see that button on the top. Third one from the left? Thats called the search button. It is your friend. Also google is your friend too.
 
In short, the first number is usually the most important... it's the "CAS Latency" rating... without getting overly technical, I'll dumb it down for simplicity sake; it's a rating of how many cycles the memory waits to make another calculation... it's faster to wait 2 cycles then wait 3 cycles. Of course a cycle isn't exactly like 5-minutes a cycle and you may or may not notice the differance, but there IS a differance, and CAS Latency is more important on P4 systems then AMD systems.
 
Originally posted by: Wolfshanze
In short, the first number is usually the most important... it's the "CAS Latency" rating... without getting overly technical, I'll dumb it down for simplicity sake; it's a rating of how many cycles the memory waits to make another calculation... it's faster to wait 2 cycles then wait 3 cycles. Of course a cycle isn't exactly like 5-minutes a cycle and you may or may not notice the differance, but there IS a differance, and CAS Latency is more important on P4 systems then AMD systems.

I thought that latencies were a lot more important on an A64 system then a P4. I thought that A64's like low lantency and P4's like lots of bandwith.

But your argument still stands up. For example:
My friend has DDR400 RAM at uber tight timings of 2-2-2-5. I have DDR550 RAM running at 3-3-2-8ish. Now, whos RAM will win 100% of the time? Mine will. It really depends on the RAM and timings. I can push my RAM to uber high frequencies at not so high latencies, so for me it's a win win situation. Most people would have to relaxe them to get to such high speed.
 
Originally posted by: wizboy11
Holy sh!t, I used google and it was the first entry:
http://www.andrew.cmu.edu/user/fma/P4tweakRAM2.htm

Even though it has to do with OCing ram on a P4, it gives an explination of everything. I got another link somewere about ram timins, I'll see if I can't find it. And you see that button on the top. Third one from the left? Thats called the search button. It is your friend. Also google is your friend too.

thanks a lot for your link wizboy...about the search function, there's a lot of stuff and for google, it was late at night so i didn't bother using it. you're right though i should've at least tried (especially since your first link was so useful).

so is there any general speed/timing thing someone should go for? for instance, is there a general formula to calculate RAM at x-x-x-x timing at ___MHz is equal to RAM at y-y-y-y timing at ___MHz? or is it really all by benching to find out? thanks for your help
 
thanks a lot for your link wizboy...about the search function, there's a lot of stuff and for google, it was late at night so i didn't bother using it. you're right though i should've at least tried (especially since your first link was so useful).

so is there any general speed/timing thing someone should go for? for instance, is there a general formula to calculate RAM at x-x-x-x timing at ___MHz is equal to RAM at y-y-y-y timing at ___MHz? or is it really all by benching to find out? thanks for your help
 
Originally posted by: MustISO
Ever hear of Google or the search function. This has been asked and answered many times.
Be nice. I'm sure every question asked in the Anandtech forums could be answered using Google, but people post because they can learn even more about the answer to their question in a discussion format. Sometimes people want to know about something rather than just the straight answer. Those people constitute a large percentage of AT members and forum-goers in general.

Still, I do agree with you about using search engines, whether Google or the AT search function. I never create a new thread to ask a question without spending at least 5 minutes Googling or searching the forums here first.

 
Originally posted by: tyborg
it depends. I find higher megahertz means a better performance gain in general. it does get to a point when bad timings ruin it though. You can look at various benchmarks AT has ran. Take a look, for example, at some of the graphs on the latest 1GB Ram article: http://www.anandtech.com/memory/showdoc.aspx?i=2560

i read this article and the 3 tested were not the fastest modules compared to the other ones. i know the others were 2x512mb kits and not 2x1024, but does that really make much of a difference?

one last thing. i know with cpus, the week you get your cpu or something like that determines how good its OCing abilities are. for ram, if two people have the same exact setup, will the ram be able to perform at the same peak performance?
 
steppings matter for cpus, right? so if you have a good stepping you get a good overclock. is there a similar thing for RAM?

also, if computer "A" can overclock to ddr533 with the exact same parts as computer "B", can computer "B"'s ram hit ddr533 with the exact same voltage and timings? EX. computer A's ram can hit ddr533 at 2.6V with 3-4-3-8 timings. does that mean computer B can reach ddr533 speeds at 2.6V with 3-4-3-8 timings?
 
Back
Top