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

memory timing question

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

medo145

Member
Joined
Jun 11, 2004
what timings are better 2-2-2-5 or 2-2-2-6? i recall someone telling me that 2-2-2-6 is better for some reasong. i just wanted to double check. i can run either one of them with no problem.
 
RedDragonXXX said:
Tighter timing are always better.
I disagree. Increasing tRAS by one might not affect performance at all. For real-world aps, even tRAS 8 keeps up with tRAS 5. But if tRAS is too tight, performance may suffer even if the system appears stable. Mushkin explains it better than I can this late at night.
http://www.mushkin.com/doc/techSupport/papers/latency.asp

I'd bench them both and go with 6 if they bench the same. I'd also give it a hard run on Prime95 before using 2-2-2-5 as my 24/7 timings. If I didn't have time to bench it, I'd just use 6, as setting tRAS lower than tRCD+CAS+2 is asking for trouble.
 
Otter said:
I disagree. Increasing tRAS by one might not affect performance at all. For real-world aps, even tRAS 8 keeps up with tRAS 5. But if tRAS is too tight, performance may suffer even if the system appears stable. Mushkin explains it better than I can this late at night.
http://www.mushkin.com/doc/techSupport/papers/latency.asp

I'd bench them both and go with 6 if they bench the same. I'd also give it a hard run on Prime95 before using 2-2-2-5 as my 24/7 timings. If I didn't have time to bench it, I'd just use 6, as setting tRAS lower than tRCD+CAS+2 is asking for trouble.
I liked the "book opening and closing in your face" analogy. :)
 
IWasHungry said:
I liked the "book opening and closing in your face" analogy. :)
Yeah, that's a good way of explaining it. What I've never understood is why slamming the book in the CPU's face can slow things down without crashing the system or at least causing major stability problems. Perhaps the memory controller is smart enough to try again when that happens.
 
Back