Alot more to performance than pipelines. All that effects is how much you can do at once, which was obviously what Core was about. Bottom line is one instruction only runs the same response time through shared power logic on teh 1.8 as she does the 3.6. No thx to you overclocking nuts that don't recognize teh glass ceiling.
So, clock for clock, an AMD Athlon based processor > Netburst based processor.
That math doesn't work out well if AMD is 20-40% under clockspeed, imo.
(why the top core proc runs at 2.4Ghz instead of the 3.6/3.4 ect you see in P4s).
2.93 Core already on shelves with 3.2 due in Janurary in the form of an EE. Awfully quick releases dontcha think?
Core also features intelligent shared cache, as well as a quantity of floating point units which make it so adept at rendering and such.
Core doesn't share a floating point scheduler liek teh AMD. That's K8L playing that game. Last I heard about K8L, it's only advantage will be running heavy floating-point SSE trash. It's going to get cleaned like a clock on the Integer side. Actually a huge drawback right now with K8L not being able to retire 4 commands per clock like it was spec-ed to do. All it's going to do is eliminate K8's current bottleneck...AMD is screwing around in too many markets now for their current size so I don't forsee K8L hitting the market fast enough anyway...Intel will be up to a 256 core processor by 2015 supposedly.
Seen the new shiny Itanium? True 64 bit lives again. I think my days of brand loyality in all aspects of the computer industry is quickly coming to an end...