Differing Memory Controllers - 3200+/3400+COCH

[DaGoof]

I recently had the opportunity to trade my aging 3200+ CH CO for a slightly less antiquated 3400+ CH CO. My old 3200+ hit its limit at 2.45ghz, and this new 3400+ tops out at nearly 2.65ghz. Although they're not record breaking clocks, the thing that struck me was the difference when pushing my OCZ 3200 EL Plat. Rev2 to it's limit. On the 3200+ I only struck 300mhz on one occasion and barely ran superpi before BSOD'ing, whereas on the 3400+ recently topped out at 323mhz, and was able to not bluescreen [despite being atrociously unstable]. Not a small difference by any means, with a differing bandwidth of 357mb/s in sandra with identical latencies. Hoping to move up to a winchester in the near future, I can only hope to see the same type of increase. The difference of this size was pretty impressive to me, considering they're both CO's, as I've never been blessed with the chance to own a CG.

http://dagoof.no-ip.com/oc/xsram1.gif

http://dagoof.no-ip.com/oc/xsram2.gif

[IvanAndreevich]

That's wicked. Are you sure the only thing that changed is the CPU?

[Revv23]

nice increase...

i thought the memory controllers were more consistent then that...

[DaGoof]

Just different cpus, only about a month difference in dates, also.

[reject]

nice, but does 3dmark change at all?

[DaGoof]

Not stable enough for 3dmark at that speed - but it can handle 3dmark at higher speeds than before

results are not comparable because the processor speeds are different, but yes, it's a higher 3dmark.