What factors help make RAM more stable

[Melchie]

ok i've been doing tons of testing at many different FSB's im using ddr 466 corsair XMS memory on a AN35N Ultra, at FSB 225x10@1.5v memory cas 3 - 3 - 3 - 8 (auto) and memtest was stable , error free after about 12 hours of total testing. prime was error free after about 4 hours testing.

i recently bumped up to fsb 460 (230x10@1.6v) and after 6 hours of memtest i got 2 errors, memory timing is cas 3 - 3 - 3 - 8 (auto)
will changing the 8 to say 11 make it more stable? or does more vcore play a role in this ? i just want a rock solid system with the 460 fsb, anything less is kind of defeating the purpose of the 466 rating on my ram IMO. thanks melchie

[johan851]

More vcore can play a role, depending on your CPU. It's a possiblity worth looking into, at any rate.

If I were you, I would try to tighten my timings. At least try to get down to CAS 2.5 or so. Also, increase your memory voltage if you can. 2.6v or 2.7v is perfectly safe, and it'll help maintain stability.

A tRAS setting of 11 over 8 would probably be a little more stable AND increase performance. I would do it.

[flapperhead]

it is very important to test each component seperately first.that way u know what is failing. if u run both the memory and cpu without knowing either ones limit,u are stuck messing with both ur memory/cpu/ and sometime the mb when something fails.

[Melchie]

ok well i decided to change my vdimm to 2.7 instead of auto, and i upped my vcore from 1.6v to 1.625. i kept the memory timings the same for now, just to see if i got the same errors in memtest and so far im looking good. 8 hours and no errors as of yet. thanks for the suggestions.