That's not accurate - when G4s were competing in the market against PIIIs/Athlons, the efficiency of scalar code (ie not Altivec or SSE) was comparable. Lots of apps (eg plugins) did not use Altivec so plugin instances were comparable at similar clock speeds.griels wrote: Maybe they're overstating their case a little, but it should be very usable. You are aware that an 800MhZ Mac is much faster than an 800MhZ PC?
P4s were (and still are) less efficient per clock but with Northwood they were the best game in town - fastest memory performance and with such high clockspeeds they outran everything G4 (by a mile) G5 and Athlons until more recently when AMD has bridged the gap again.
Lets hope that they are in fact overstating the case here. It is hard to see how Altivec optimised code could be ported on the fly to SSE so I expect Logic is not a good contender for this sort of thing.
Eg
Of course, it's a suboptimal solution, and I hope it prompts Apple to wise up and provide a better one (OS X on x86 and an accompanying version of Logic), and heats up the competition on the PC side for sequencers of comparable sophistication.