[Ti] A real world comparison

Kynan Shook kshook at cae.wisc.edu
Thu Jun 9 13:17:36 PDT 2005



Chris Olson <chris.olson at astcomm.net> writes:
> You can easily heat a small room in your
> house with the P-IV processor that Apple is shipping in the dev boxes,
> for instance.

I thought the same thing too, until I felt the air coming out the  
back of the machine.  It's much cooler on the developer machine than  
a current PowerMac.

> This is exactly what happened with the G5 getting to 3.0 GHz.  It's  
> all
> marketing.  Had nothing whatsoever to do with practicality.

The point isn't that the number failed to meet the target - the point  
is that, no matter how it is done, performance hasn't increased as  
much as it needs to in order to keep up with the state of the art.   
And, Apple has determined that this will continue to be the case -  
Intel's speed increases are expected to be greater than IBM's.  If it  
were all about the clock speed, Apple would have switched to Intel a  
long, long time ago.

I'm very happy with my new Dual 2.7 GHz PowerMac.  It's a great  
machine.  But it sure makes my bedroom a LOT warmer.

Why don't you consider adopting the wait-and-see attitude that so  
many others have adopted?  An Intel Mac will probably beat a PowerPC  
Mac at many tasks, and a PowerPC Mac will probably beat the Intel Mac  
at many tasks.  But as time wears on, the Intel Mac will improve more  
rapidly.

Does anybody else recall there being nearly this much uproar when  
Apple has made changes within the PowerPC architecture that had  
actual numbers showing performance differences (both good and bad),  
instead of just all this theory and FUD that has been thrown around  
recently?
For example: when the G4's pipeline was lengthened from 4 to 7  
stages.  Or, similarly, the change from a 1 MB L2 cache at half the  
core speed to a 256 KB L2 cache at full core speed.  For that matter,  
the G5 PowerMacs weren't always faster than the G4 PowerMacs before  
them.  All of these changes turned out to be good in the long term,  
even though some benchmarks might have taken a step back.  Apple  
won't ship a system that's slower overall just because it has an  
Intel chip.  A few metrics might decrease a little bit, but most will  
increase significantly.  If they don't, then you can bet the PowerPC  
has a longer life in Macs than currently planned.


More information about the Titanium mailing list