Originally posted by: Vehementi
Originally posted by: butch84
So, why cant you just install osx on a pc??
OS X Jaguar is lurking around out there. Run a search for it
All I know is that the Mac architecture is RISC, which is why a 1GHz G4 can rival a Athlon XP 2000+, in MFLOPs and MIPs.
Adapting to the Macintosh Architecture
If any of you know where a comparison between, say, a G4 and an Athlon XP or a G4 and a P4 is, let me know
That's not very true at all. Here's a popular example, it's a little old, but it's somewhat relevant:
http://www.digitalvideoediting.com/2002/05_may/features/cw_aeshowdown.htm
Basically it shows a dual Athlon MP (1.533GHz/1800+) against a dual 1Ghz G4+ system. The Athlon system is clocked 50% higher than the G4+ system and beats it on average (not losing a single benchmark, leading anywhere from 30%-50%) of ~40%. And this is on After Effects, Adobe, Mac's home turf. Adobe isn't exactly a Carmack of optimizing x86 code.
Mac's greatest strengths do not lie in brute force (because even the latest dual 1.25GHz G4's pale in comparison to the highest end PC stuff), it lies in stuff that PC's just can't have (and it's not just OSX). One of those things, one BIG thing, is color calibration. Mac's color calibration is just
so far ahead of PC's that it's sickening. Actually what's sickens me is the incredibly poor color calibration of PC's. Not only are Mac's color calibration painless and seamless between
all its programs and hardware, but its hardware and operating system are symbiotic with it. Their monitors send info to the system, the system sends info to the OS, and back again. It's so sophisticated that as an Apple monitor ages and some of its properties degrade, the OS will detect it and tell the monitor to adjust its properties accordingly, so it will keep displaying correctly until it's so far gone you need to chuck it out.
I suppose at the most basic level, you can boil down the difference between the PPC architecture and x86 architecture (and subsequently why it's inherently incompatible) to its operand system.
x86 uses the two operand system, ie. a + b = a. Or in English, two variables, 'a' and 'b', are added together, the result is recorded (written over) the variable 'a'.
PPC uses the three operand system, ie. a + b = c. Although with the introduction of AltiVec, the G4+'s actually use four operands in SIMD operations. It gets kind of complex.
This reason alone makes it VERY, VERY hard to compare performance between one and the other. The above link is one of the best, nearly unbiased comparison, and it's still far from perfect. The best comparisons you can hope for are "real world" comparisons like FPS benches or application timing (like the After Effects benchmarks), but even that's not entirely reliable because very few companies have the talent, time, or resources to optimize both sides of the coin equally. Microsoft, suprisingly, is probably the best at it. If you use Office in OSX or Office in WinXP, they're both pretty much equally fast. IE is still the fastest browser on either platform. Go figure.
The differences in operand systems and how they affect performance is also hard to predict. It's pretty obvious that the three operand system is better, although the why is not quite what you'd expect. Performance-wise, they're almost equal, but x86 has to do some compiler and scheduling tricks, luckily for x86, Intel is the master at this. The biggest advantage of the three operand system is flexibility and ease of use.
Going back to your first question. It'd be like trying to put three numbers in a two-variable equation or vice-versa, ie. the square peg in the round hole.
Facebook
Twitter