SSE4 is what is most concerning to me. The fact that Q6600 does not have that, and Q9xx does. In fact, for that matter, if the SSE4 can make that noticeable difference towards encoding, then the Q9300 isn't ruled out either. Considering the chip is just $50 more then Q6600.
From what I gather, on my P35 setup, on air, I can OC the Q6600 to a stable 3.4GHz-3.5GHz, and the Q9300 to 3.2GHz. (Correct me if I'm wrong. That being the case, the 0.2-0.4GHz is most likely not going to be a HUGE difference. However, in case of video encoding, the SSE4 instructions "might" make the deciding difference.
There is no denying that the Q6600 will OC a lot more then the Q9300 (Or Q9450) for that matter, but it will lack the SSE4 which may impact Video Encoding quite a bit more then the 0.2-0.4GHz (Again, correct me if I'm wrong on this).
Quite honestly, this thread should be re-titled "SSE4 - Worth $50". I was considering the Q9450 earlier, but I don't thing I'll invest that much in the 45nm/Yorkfield for now - It'll either be Q6600 or Q9300 for now. I'll wait for Nehalem or even Sandy Bridge in about 2 years or so to replace whatever I buy now.
But anyhow, now to determine between Q6600 or Q9300 - it all depends on approximately how much more performance/speed I can gain with the SSE4 - Which brings me to your speed-test offer.
Can you possibly use the latest TMPGEnc Xpress (Which supports SSE4, post November, 2007 release) and encode an h.264 (MKV, MP4 container) to a WMV-HD (VC-1) Advanced Profile video and see if the SSE4 makes that drastic difference?
I'll appreciate if you can get back to me on that.
Thanks a ton!




Reply With Quote

Bookmarks