Motorola Unveils Chattier Chip

[sheepguy42]

Originally posted by azosx

Nobody optimizes their apps for AltiVec. Ok, you're right, Adobe does, my bad. Apple was hell bent on creating this specilized processor much like the SPARC or MIPS. Highly optimized, low core clock speed and very expensive to produce. Unfortunately, the only people who did take advantage of this was Adobe with Photoshop. Everybody else writes to it like it has MHz to spare and that's why every windows port runs like crap and everything else is just feels a little slow and out of place.

I would just like to see Apple succeed this decade and sitting on their hands, waiting for some POS CPU isn't really a good start.

I mentioned this earlier, but as SimX likes real evidence (and maybe azosx will notice too) here is the link: http://www.apple.com/macosx/jaguar/unix.html
and here is the quote:

Originally on Apple-Mac OS X-Features-UNIX

Developers will appreciate the inclusion of GCC 3.1, which offers 25% better code generation and compiles your code in as little as one sixth the time. Also, Apple engineers have optimized the standard suite of math libraries to use the vector instructions on the PowerPC G4 chip (also known as the Velocity Engine). That means you can get optimal performance for your applications without writing assembly or platform-specific code.

Now, as far as I know, this means more apps should start using AltiVec without even meaning to, and performance on all apps should improve quite a bit with future versions (made with 10.2, for 10.2). That's the best I can offer.

 

[phatsharpie]

Originally posted by azosx


They're not shipping because Apple has no intention on releasing a 1GHz G3 next to their 1GHz G4. IBM doesn't have yield problems, Motorola does. Because of this the G3 is stuck in limbo.

IBM released information about their 1GHz G3 almost a year ago. Are you telling me they haven't figured out how to make it run yet? If not, then Apple was pretty stupid for going with the G3 or G4 considering how long it takes to increase the MHz.

Common sense says IBM could make one hell of a processor for Apple. They certainly have no problem putting higher clock speed G3 variants into their highend mainframes. IBM said the hell with AltiVec a long time ago and decided high speed, low power consumption was the answer. Apple didn't bite and now we're stuck in slowville.

Don't get me wrong, my G4 667 does exactly what I need it for and at reasonable speeds but is truly a dinosaur next to my PC.

Let's get one thing straight here... Releasing information about a processor and actually shipping the processor in volume are two very different things. If I remember correctly, Motorola was demonstrating 1GHz G4s at microprocessor trade shows WAY WAY before the chips were ready to be shipping in volume. In essence, it doesn't matter when IBM released information about their 1GHz G3s - until they ship it in volume, it means squat.

Also, IBM does not use the G3s in their mainframes. They don't even use the G3s in their UNIX workstations. This is simply because the G3s do not handle SMP well. I am assuming that when you say IBM uses a "derivative of G3" in their mainframes, you are referring to the POWER architecture. To say that POWER is a derivative of the G3 is a bit naive. The POWER series of chips have been around far longer than PowerPC. And to state that the POWER-III architecture is a derivative of the G3 is like saying that a Porsche is a derivative of the Volkswagon. Its true that they have similar heritage, but they are vastly different. The POWER chips are aimed at the mainframe, UNIX workstation, supercomputing market. They are powerful but are resource hogs and EXPENSIVE. However, they can indeed run PowerPC code. This is because the origianl PowerPC design was based on the POWER-I.

All the above information can be confirmed at the IBM website. I think that if anyone want to make an argument that IBM is a better way to go for Apple, that is fine, but please do some research and not base your arguments on hearsay and rumors.