Originally posted by: Brahmzy
lopri,
what do you think of the P5W DH?
I have a retail E6600 that I got to 3.2GHz @ 1.45V stable and 3.3GHz@ 1.5V stable on a DS3.
Well, I blew up the DS3 due to a bad BETA BIOS flash mishap and I ordered the P5W DH. However, the more I read about it, the more I see it is plagued with problems-extremely slow boot times being one (which would REALLY piss me off on a $250 board. Should I have gotten another DS3?
I also have a Thermalright Ultra-120 - I LOVE it. People bash it, but I think they're high.
Brahmzy:
I'm quite cautious to even begin speculating on this, but the more and more I play with E6300 and E6600, I get a weird feeling. When I put my E6600 on P5W-DH, its behavior is quite erratic. Can't example those here but the difference is definitely there. Once I put the E6300 in, all the undesirable behaviors are pretty much gone and the performance becomes once again predictable.
1. (Very very carefully) It could be Intel that messed up with Rev.B Conroe?
2. Chipset handling 2MB L2 Cache and 4MB L2 Cache differently? (more likely)
Coming from Socket 939, this might sound like a BS, (since Toledo/Manchester, San Diego/Venice are just that, half the cache, and we didn't see any difference other than clock and power/heat. Or we could even say that the L2 cache size was transparent to the chipset.) But from my observation so far, there is definitely something else going on between Conroe and Allendale. Not only the L2 cache size, be it the chip's architecture or the way it communicates to the chipset (or the way chipset treats 2MB L2 vs 4MB L2).
Hope someone will eventually let us know what's going on..
For now, my first recommendation for E6600/E6700 is still P5W-DH. I haven't had any P965 board yet and from what I've heard they're still immature compared to 975X.
Sorry it wasn't really a clear-cut answer. Oh and I was dozing on my desk. zzz..
Good night all