Originally posted by: skyeclad... latency issue to make an upgrade from 1GB(2x512) to 2GB(4x512) using OCZ PC3200 Premier RAM worse than not upgrading at all?
Do the 4x512.
I started out building my son's PC using 2X512 OCZ "Gold" versions, and by sheer coincidence, each pair had a bad module. I never even got a chance to OC the memory as a pair. As NewEgg took care of me each time, I just kept "upgrading."
OCZ4001024WV3DC-K
OCZ4001024ELDCGE-K
My son intended to play BF2, and I read Zebo's Memory Latency article enough times that I just gave up on trying for a low-latency OC, and "upgraded" to 4x512 Corsair VS. No regrets. Just plugged it in, set to 2T (and Bank Interleave 'off'), and it was done. (I got tired of OC testing, and the Gold gave off a lot of heat for the small room.)
I recently tried VS 2x512 @1T (& bank interleave on) versus VS 4x512 @2T (& bank interleave off), everything else the same. 2x512 1T was 0.2% faster in 3DMark05 test results on my system. My son played BF2 before I put it back to 4x512. BF2-Online was stuttering and 'lagging'. After 4x512, everything was back to normal, smooth play. ( I ran Task Manager/Performance and made him restart BF2 while 2x512. CPU was pegged. I assume it was swapping page files.)
A younger son is soon to get a system upgrade, so some OCZ Platinum 2x1Gb is on the way. The Platinum is $30 less (AR) than what I paid for the VS on sale 9 months ago. I will try some bmarks to see the difference between 2x1Gb 2-3-2-6 versus 4x512 2.5-3-3-7 at various speeds in the same system.
What is a good bench mark to use for performance testing? All I have is the freeware 3DMark05 tool. Is there a 'standard' BF2 benchmark that does not require a 3rd-party FPS measure?