G.Skill F1-4000USU2-2GBHZ DDR low latency?

[Lord Banshee]

Ok willing to lose bandwidth if anyone has any tricks to make this ram run lower latency? I am not sure i a bought it instead of the low latency one ... maybe a year ago i want more bandwidth lol

my current settings at stock voltage are in my sig..

Thanks

 

[Nickel020]

Those are based on Samsung UCCC chips, you might be able to get 3-4-3 but that's about it. Even at DDR400 these won't get you low latencies. Infineo chips give lower latencies, but often don't overclock as high as as UCCCs.

And what's with the low latency anyway? Even if you were to run 3-3-2 or something at 240MHz, that would be a noticeable performance loss.

In German, but a good reference anyway:

http://www.forumdeluxx.de/forum/showthread.php?t=177162

http://www.forumdeluxx.de/forum/showthread.php?t=186138

 

[gOJDO]

Something confuses me. If your Opteron is working at 289x9, then how your RAM works at 260MHz?
It might be 240MHz or 289MHz(480MHz or 578MHz DDR).
I had Athlon64 3200+(Venice) @2.8GHz(10x280) with A-DATA DDR-500 PC4000 CL3 which worked at 233MHz(466MHz DDR) CL2 3-3-6 CR 1T, at +0.2v
I think if you are running your RAM at 240MHz, then there would be no problems for it to do the same.

 

[Lord Banshee]

Nickel020 i just wanted to see if less latency is better than more bandwidth... I'll check out the link.

gOJD0, dividers. (180/200) * 289 ~ 260

 

[gOJDO]

Well, that makes a sense. My mainboard, Asus A8N5X, was supporting 5:6, but not 9:10 memory divider.
I made a comparison between the performance of 560MHz CL3 4-4-7 CR1 and 466MHz CL2 3-3-6 CR1, and I concluded that with the second settings my system was faster for everything. Even my Sandra Memory Bandwidth score was higher. Anyway the performance difference was little, 2% in average.