• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Memory Timings and MHZ Performance Explained

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Time4aMassiveOC

Member
Joined
Dec 10, 2004
Location
The CircuitCity FireDog House
heres the link to the info

http://www.hardwaresecrets.com/article/167/4


to highlight the important parts
as we all seem to know Cas (the first timing number) is by far the most important in terms of memory performance. as highlighted here


"If we compare a DDR memory to a DDR2 memory running under the same clock, the one with the lower latency will be the fastest. Thus, if you have a DDR400 CL3 memory and a DDR2-400 CL4 memory, your DDR400 will be faster."


however what isnt common knowledge is this


"Keep in mind that DDR2 memories have an additional parameter called AL (additional latency), which must be added to their nominal latency (CL) in order to get the total latency."


so this means

(CAS + AL) x (Clock Tick Duration) = Main Performance for DDR2

vs

(CAS) x (Clock Tick Duration) = Main Performance for DDR


"The additional latency (AL) is usually found on the memory module technical specs (usually a PDF file for downloading on the manufacturer website)."


now the breakdown of MHZ and how it works with the timings.


"On a DDR400 CL3 memory, this “3” means that the memory delays three clock cycles to start delivering the requested data. Since this memory runs at 200 MHz, each clock tick measures 5 ns (T= 1/f). Thus its latency if of 15 ns.

Now on a DDR2-533 CL3 AL0 memory, this “3” also means that the memory delays three clock cycles to start delivering the request data, but since this memory runs at 266 MHz, each clock tick measures 3.75 ns, so its latency is of 11.25 ns, making this memory faster to data delivery than our DDR400 CL3 memory. So a DDR2-533 CL4 and AL0 memory has the same latency as a DDR400 CL3. Notice that we are assuming the additional latency as zero, or we would need to take it into account, i.e., a DDR2 CL3 AL1 memory has in reality a latency of four clock cycles which would equal 15ns."


the clock tick duration for memory speeds

Memory.........................Clock Tick Duration (each one)
DDR266.........................7.5 ns
DDR333.........................6 ns
DDR400 and DDR2-400.....5 ns
DDR2-533......................3.75 ns
DDR2-667......................3 ns
DDR2-800......................2.5 ns






dont get AL (additional latency) mixed up with the other timings. the other timings mean this

* CL: CAS Latency. The time it takes between a command having been sent to the memory and when it begins to reply to it. It is the time it takes between the processor asking for some data from the memory and it returning it.
* tRCD: RAS to CAS Delay. The time it takes between the activation of the line (RAS) and the column (CAS) where the data are stored in the matrix.
* tRP: RAS Precharge. The time it takes between disabling the access to a line of data and the begin of the access the another line of data.
* tRAS: Active to Precharge Delay. How long the memory has to wait until the next access to the memory can be initiated.
* CMD: Command Rate. The time it takes between the memory chip having been activated and when the first command may be sent to the memory. Sometimes this value is not informed. It usually is T1 (1 clock speed) or T2 (2 clock speeds).



ENJOY :)
 
Last edited:
assuming the DDR2 actually had zero additional latency (AL)


CAS 2 @ DDR540 = about 7.5ns
CAS 3 @ DDR2-680 = about 9ns

and thats assuming that the DDR2 doesnt have AL!
add one clock cycle to that DDR2 and now you are up to 12ns!

makes ya wish intel had CAS 1.5 DDR533 for thier 1066FSB processors!

Edit: changed equation to more realistic cas settings
 
Last edited:
_damien_ said:
CAS 1.5 doesn't exist for any memory architecture - this would've required an additional pipeline stage, and can't be helped by fancy BIOS-options or memory controllers.
Also, for DDR2, half CAS values don't exist.


i was just replying to a guy over at XS when he said

"whats the balance?? - UTT / BH5 1.5(2)-2-2-5 @ 270 beated 2.5-3-3-x @ 300 mhz, and 2.5-4-4-x @ 340mhz?

Is 3-2-2-8 in DDR2 at 300 mhz faster than 5-5-5-16 @ 600mhz???"

i agree with you on the DDR2 whole numbers, i wasnt aware that DDR2 even goes below CAS 3 i was just doing the math for the question he asked.

as to disputing the definitions of the timings, i just copy pasted them directly from the site

im guessing they were shooting for people grasping the simple idea of what each was. not explaining it to a memory expert. but thats just my guess
apparently it was too laymens for your taste :) i appreciate the extra detail


could you point us/me to a site where there are better definitions for all the timings you were talking about?

but you arnt disputing the fact that CAS timing is the most influential to system performance.
 
Time4aMassiveOC said:
CAS 2 @ DDR540 = about 7.5ns
CAS 3 @ DDR2-680 = about 9ns
Why would you compare highly clocked DDR to practically stock setting DDR2? Take Micron D9 (3.7ns) bump it to CAS 3 @ DDR800 and it'll give 7.5ns - also, b/w on DDR2 will be huge compared to CAS 2 @ DDR540. :-/
 
as far as i know the gains of running ram at speeds that exceed 1:1 with the processor give incredibly small performance increases.

as to why that is im not sure. ive always been of the opinion that its due to the memory having more bandwidth that the processor does. whereas at 1:1 the memory bandwidth and processor bandwidth are identical.


so the reason i used those speeds was because i was quoting the guy like i said in the above post. however, if you look at the speeds not many people can run DDR2-680 at 1:1 , in fact you can probly count them on your fingers.


like when crimedog was running his ram redlined :drool:

343FSB 1:1 DDR2-686 @ 4-2-2-8 = 8378 : 8785 bandwidth score

312FSB 3:5 DDR2-1041 @ 5-3-3-8 = 7970 : 7969 bandwidth score

oh and for comparisons sake

CAS 2 @ DDR400 = 10 ms

CAS 4 @ DDR2-686 = 11.9 ms

and thats if he had no (AL) but his probly had at least 1

CAS 4 + AL 1 @ DDR2-686 = ~14.85 ms

which is probly why he said this in reference to his latency score in everest of 67ms vs the stock DDR400 score of 45ms

"i'm not saying that ddr2 is faster, i'm not saying it's slower. i just think that benchmark is flawed because i'm running ddr2 really-really fast and it's slaughtered by ddr1 @ stock. "


heres the link to the post if you wanna see pics of his badass scores among others

http://www.ocforums.com/showthread.php?t=419128&page=1&pp=30
 
Last edited:
Back