Hardware questions!

By RadishTM
in Hardware Hangout

 Share

Recommended Posts

Grand Master

RadishTM Veteran

Posted
  • Author
  • Veteran
Posted
I never use Everest myself (maybe it list the timings different than what I am used too), mind posting a screenshot of the CPU-Z memory tab?

P.S. Nice Sig/Avatar

Thanks, here you go -

Radish?

post-4583-1184868453.jpg

Mentor

12Iceman

Posted
Posted

Yeah, the way timings are most commonly listed you are running at 5-6-6-18. I doubt that running at 5-5-5-15 will give you any noticeable increase in performance (still CAS5), but you should have no problems running at those timings since your RAM is rated to do so, and theoretically it would only help you. If you want to get a recordable peformance increase (still probably only like 3%) you can try running your RAM at 4-4-4-12 which should bump your SuperPIMod and 3DMark06 scores up a tad bit, but there is no guarantees your RAM would be stable with those timings so you would have to test it.

Grand Master

Jason S. Global Moderator

Posted
  • Global Moderator
Posted

5-5-5 for 667MHz is really loose. i forget which ram you bought though. you could most likely tighten those up to 4-4-4 or possibly 3-3-3. im running mine 740MHz @ 3-3-3-8.

the difference b/w 5-6-6 and 5-5-5 would be nil... perhaps a few extra points in 3dmarkxx, but other than that you wont actually notice any speed increase.

play around w/ your timings... you might be surprised.

Grand Master

gwai lo

Posted
Posted

It's command rate and can offer 10-20% in memory performance, but in the real world you're looking at less than 3%.

Well, in short, it allows the memory controller to send out a command every cycle ( with 1T ) and every two cycles with 2T. Thus, obviously the first is harder on the memory controller thus may affect your overclock's stability.

  • 2 weeks later...
  • 4 weeks later...
This topic is now closed to further replies.
 Share
Go to topic listing

  • Recently Browsing   0 members

    • No registered users viewing this page.