[siliconwafer]

http://www.haxial.com/spls-soapbox/apple-powermac-G5/

It's a great article which sums up a few things: Apple crippled the x86 chips and optimized for the G5 in their own benchmarks. (disabled hyperthreading, didn't use SSE2, etc.)

I now feel a bit betrayed by Apple, as these benchmarks clearly show that the G5 is slower than the Intel P4. Let alone, his comparisons don't even include the fastest P4 on the market right now.

[Spliffdaddy]

I noticed this was posted on the MacNN homepage last night, but now it's mysteriously been removed.

[InterfaceGuy]

There is already another thread discussing this.

http://forums.macnn.com/showthread.p...hreadid=165758

[The Placid Casual]

Apple didn't cripple x86 anything. Period.

They used gcc3.3 as the compiler, a 'real world' compiler much more relevent than any other out there...

They just did not use the uber optimised Intel made compilers which are to be honest crooked... they tilt things even more in favour of the x86 architecture, and give very unfair results...

Please note that gcc3.3 IS still heavily optimised for x86 anyway... however most apps use this, and not the specially made Intel ones...

The guy who wrote the article, was 1. A troll, 2. Very uninformed...

Why was he going on about Integer Specs when we know all the Mac apps rely on decent Fp results? The price difference of $1. No 64 Bit OS? WTF? The article was a PC troll joke.

The Mac is back in business, please don't feed the trolls crawling out of the woodwork...

[Scotttheking]

http://forums.macnn.com/showthread.p...hreadid=165758