6-3-3-2 vs 7-3-3-2 corruption theory?

Status
Not open for further replies.

DemonEdge

Baseband Member
Messages
38
I've read in a statement from Corsair that:
"tRAS = tRCD + CAS + 2 (lower than this will result in data corruption and crashes)"
& everytime i go check in a review of a cpu/mobo/ram i find these reviewers violating this very rule
they usually go 5-2-2-2 instead of 6-2-2-2
or 6-3-3-2 instead of 7-3-3-2
Now I ask myself...are the people of Corsair lieing or are the reviewers ignorant?
Can anyone help me solve this problem?
 
Depends on the RAM. and following that rule you have with 2-3-3-6 it'd be 2-3-3-8 instead of 7, but yeah, that only applies to certain RAM sticks as the high performance RAM has the tightest timings of 2-2-2-5 and there's obviously no corruption otherwise they wouldn't be selling it.
 
Nubius said:
Depends on the RAM. and following that rule you have with 2-3-3-6 it'd be 2-3-3-8 instead of 7, but yeah, that only applies to certain RAM sticks as the high performance RAM has the tightest timings of 2-2-2-5 and there's obviously no corruption otherwise they wouldn't be selling it.
why should it be 2-3-3-8?
tRCD + CAS + 2 =tRAS
3+2+2=7... not 8 :eek: :)
and for the other topic i am telling you that CORSAIR is the one who posted that issue (if you want the exact address i can look for it) and while we still on that...mention any better ram manufacturer you might know.... ;) :p Because i don't know of anyone better than CORSAIR
 
I know that you said corsair said it, I wasn't argueing with that.

And based on the timings being 2-3-3- it'd be 2+3 = 5 + 3 = 8, so now you've changed your timings from 3-2-2-7 to make it work for equaling 7 , I was saying if it was 2-3-3- THAT would be 8.
 
you cant really say that one brand of ram is better than another, because different companies use different IC's (Integrated Circuits) and in some cases, different PCB. it all comes down to those key points, IC's and PCB. for example, a long time ago now, corsair released their 2-2-2-5 PC3200 ram on the standard PCB, and more recently, OCZ (and some other companies) have released ram with the same timings of 2-2-2-5 PC3200 with a new kind of PCB, which was made by Brainpower. Both companies use the same IC's. (Samsung TCCD) and both the corsair and OCZ ram perform the same at their stock speeds, yet, as you start to push the ram above its standard speeds, the OCZ ram is capable of reaching speeds over 30mhz over that of the corsair ram.

im not sure if that even said what i was meaning it to say, its a bit of a rant really, but you cant pitch on companies ram against anothers because the performance really comes down to what IC's the ram uses and what PCB it uses.
 
Hence me saying depends on the RAM ;)

Got that link I gave you Raekwon, cause I don't have it on this computer. You could post it so he knows what you're talking about
 
yeah the BH-5's are good chips, but generally require some high voltages to really juice them of their potential.

Also the 2x256 kits run better than the 2x512 kits as far as overclocking goes.

Do a bandwidth test Ace, and you'd be surprised at how little you lose going from 2-2-2-5 to 2.5-3-3-7 for example.

4W4K3 actually did it from 2.5-2-2-7 to 2.5-4-4-7, from the middle numbers going 2 to 4 he lost like 200mb/sec bandwidth, and from 2-3 it wasn't much.

I got mine running DDR480 at the moment with 2.5-3-3-7 timings.

Got 3655/3372 in the SiSoftware Sandra Memory Benches with that.

I know it can hit DDR250 on these timings, but I don't think my mobo wants to do it, although Mike9872 whatever those numbers were lol, got 250 working on his NF7

I'm jealous! :p
 
Status
Not open for further replies.
Back
Top Bottom