It was a less than average piece if you ask me...
I didn't post yesterday because I had a bazillion problems with the DFI. This mobo has problems with the 11 multiplier (which is the stock multy in 3500+), and also has problems with A-data DDR500 which use Hynix BT-D43 chips. It took me the first day to find these out, and today I started ocing the thing using the 10.5 multy and 1/2 divider on RAM.
By saying stable, I mean it passed 20 mins without errors in Prime95 (I didn't have time to test for more yet).
Well, to make a long story short, with stock voltage (1.4) it reached 2470.
After this nothing mentionable.... I needed to pump 1.55 to make it stable at 2520 and no more than that (also tried 2550 but prime failed in 10 mins).
I guess the stock cooler is holding me back a bit when adding voltage....but then again I would expect to reach something more like 2600 with 1.55v...
I guess it's a less than average piece, nothing worth really....
However I've heard winchesters don't like prime and are stable in frequences higher than those that prime fails....I guess I'm going to test that later... At the moment I have left it at 2520 1.55v priming and I'll let it go for more than an hour, to see if it's really stable....
Things to keep in mind is that I use stock cooler, and that at 2520 1.55v full load temp is 54c ....
Andreas, the CPU was bought by a guy in a Greek forum that can fetch some CPUs...but I don't think its any good (this CPU) so don't just go buy one
TASOS, nice too see another Greek guy here.... I remember you from pctech too...
I will post any new info if I got later...
Bottom line is, this CPU is nothing special and you are better off with an older week part...