[pwnt by pat]

I read the article here and it confused me. I was always under the impression for nf2 boards that dc async was the way to go however this article says that non-dc can achieve ~5-10mhz higher fsb with relativly no performance difference. Then it goes on to explain how there is a 1-3% difference and that's where it totaly threw me. It also went on to explain how running in sync can get about 5-10% more performance or something like that...

So what is - in general - the way to go for AMD: dc async or sc async or their sync counterparts? What's the truth about sync for AMD?

[nut581]

dc sync will yield lower maximum overclock than sc sync

assuming your memory is the limit to high fsb on dc, then dc async will also yield lower maximum overclock than sc async .. in both cases, the high fsb acquired by going sc will give you pretty much par performance level with dc on lower maximum fsb.

1-3% is nothing to sweat to begin with .. as one can assume the value as the standard deviation due to different testing procedures/environment/uncontrolled factors.

[warenb]

whats the difference between dc and sc and what do they stand for?

[Sophisticated]

DC mean Dual Channel - Running 2 Sticks of Equal Size, Brand, & Latency in every other SLOT (1 & 3) or (2 & 4)

SC means Single Channel - Running 1 or more stick of any type..

[pwnt by pat]

So if I were to take my Mushkin 3500 (rated for 216 +dual pump I think?) on my infinty (which can theoretically do 270+fsb), running in sc could yeald a higher performance? Is the only way to find out test all possible configurations stock sc, stock dc, max oc sc, and max oc dc with sandra and 3dm and other benches?

[Sophisticated]

DC shouldnt limit your OC..its the type & speed of memory that can limit your OC..thats where dividers and Vdimm come into place