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

timings dont mean squat

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

EvilCloudStrife

R. Kelly is my roll model
Joined
Feb 15, 2003
Location
Mesa, Arizona
so i am playing around with my new xblk tccd, i got them up to 279mhz (i know, they can go higher) anyways. i ran some benchmarks with them at 2.5-3-3-6 and then also at 2.5-4-4-10 and the results are exactly the same.. the bandwidth is around 7000 mb/s either way, i dont see why people make a big deal about timings.. or maybe my ram is funky or something. someone care to explain?
 
EvilCloudStrife said:
then whats the point of 2-2-2 ram when 2-3-3-8 will run all the same? just a gimmick?
Have you compared the latency in Everest? Bandwidth is speed relevant, and timing sway latency.
 
I was gonna state the latency argument as well. Also use Fraps and check in game FPS to notice any differences moreso than a theoretical bandwidth test.
 
EvilCloudStrife said:
then whats the point of 2-2-2 ram when 2-3-3-8 will run all the same? just a gimmick?
recall that you're talking about the SPD information for the ram ... info that the ram makes available to the mobo/OS for configuration. while interesting to the user, the original point of the SPD info was to permit the mobo/os to configure properly for the ram in use.
 
The point is some rigs can't achieve higher bandwidth due to restriction imposed by power supply, CPU, cooling, and motherboard/chipsets. For those that want high performance but can only achieve low bandwidth, then get the low latency memory 2-2-2. What is the difference? Higher HTT (FSB) will give you more FPS while gaming!
 
EvilCloudStrife said:
then whats the point of 2-2-2 ram when 2-3-3-8 will run all the same? just a gimmick?
What benchmarks did you run? Sandra bandwidth isn't everything, its useless really. Try some PI and you will see the difference between 2.5-3-3 and 2.5-4-4.
 
one application that might give you some proof for changing ram timings is 3DMark01SE. It's not always easy to show evidence of overall system performance but I think this benchmark would be a good start. Also if you want to try one of the pcmarks you might get a better idea.
 
There is a difference, whether you notice it or not is based on what you use your puter for

The difference between cl 2, 2.5, and 3 on my machine is very noticable

cl 2 vs cl3 on superpi 32m results in 4+ minute difference (cpu at 2.7) which is quite a difference in my book
 
My benchmark.. the wife.

She can inform me with ignorance if the computer is sluggish. Since I read this thread, an experiment was performed. I backed off from 2-3-3-5 to 3-4-4-8. I was asked if the computer had something wrong with it. I 'failed to inform' the other user of the changes. I reset it to normal timings and I got a smile. Not very scientific, but there is a noticable difference to a n00b user when you raise timings. It makes the computer responce for items stored in the RAM and makes it feel "crisp".
 
Back