Meh, we'll have a Macbook air version by then. Alas not the new Atom-3 Processor ones, but the EOL Intel ones.

I have an Atom based netbook, I'd rather have a G4.

Benchmarks were updated again. After some tweaking 2.0GHz G5 lame is now 2secs faster (using the latest lame version with altivec patch and built for G5 specifically).

Also, the graphs now include 2.5GHz G5 results from my Power Mac G5. 9 seconds for lame, 39.9 seconds for mplayer.

http://sintonen.fi/pics/lame_benchmark.png
http://sintonen.fi/pics/mplayer_benchmark.png

@Piru: if you want i can test on a bi-G5 at 2.7 GHz

@Papiosaur: leave that pleasure for me ;)

heuu, no

Quote:

Wow, the 2.5GHz G5 is TWICE(!) as fast as the X1000 in the lame test, and ~75% faster in decoding the H.264 full HD movie!

And this at a small fraction of the X1000 price, and machines are actually available (unlike the X1000). And it's not even the fastest G5 either...

It has been said before, but should be said again — The MorphOS Team for sure made the right decision when going for those cheap, widely available second hand Mac machines. That was clearly the winning move!

Last time the benchmark result improvement came from using the latest lame with G5 specific optimizations enabled. At that point I didn't even consider the possibility that the generic AltiVec (G4) optimized binary could be faster on the Power Mac G5. Interestingly it is, by several seconds.

New lame benchmark results:

2.5GHz G5: 6 seconds
2.0GHz G5: 8 seconds

Updated graph (PowerBook and Mac mini results updated as well):

http://sintonen.fi/pics/lame_benchmark.png

Piru,
Quote:

Oh man is this getting goodd!