I'm over it. Are you?

[Viro]

Originally Posted by Lt Major Burns

the G5 is a pathetic chip. real world benchmarks (not apples desperate propaganda) proved this. good riddance.

Generalizations are usually wrong. The G5 is an excellent chip, especially if you're into running very very numerically intensive code. Take a look at the FFTW speed benchmarks and see how a 2GHz G5 does against every other processor out there. A G5 running binaries compiled with GCC compares very favorably against a P4 Xeon nearly that has 40% clockspeed advantage that was running binaries compiled by Intel's heavily optimizing compiler. In double precision operations, the G5 and P4 were nearly equal with the G5 being slightly better. Moving to single precision and Altivec, the G5 absolutely blows the P4 and SSE out of the water. If the G5's used IBMs own compiler, which is reported to be much better than GCC, you can expect even better performance. If you think FFTs are merely synthetic benchmarks with no applications, do a Google search.

The G5s are fantastic chips. Who cares if they can't hit 3 GHz? The only reason the P4 can reach 3+ GHz is because of it's architecture. The Athlons/Opterons which share more similarity with the G5s are struggling to go past 2.5 GHz as well. If Steve Jobs really dropped the G5 because he feels IBM dropped the ball, my opinion of him will seriously drop. This will definitely be a case of smart people letting ego get in the way causing them to make stupid decisions.

I thought Apple was trying to break into the scientific and high performance computing segments. Witness their trumpeting of the Virginia tech cluster. The G5s with their incredible Altivec vector units were a God send. The G4s had good vector units but were held back by the poor FSB. The G5s with their GHz+ FSB largely made up for this short coming.

Part of my understanding for the switch to x86 is because there is little hope of getting a G5 into a laptop. This couple with the fact that Intel has the lovely Centrino makes a switch look very attractive from a laptop POV. However, the new G4 processors from Freescale (ex Motorola) have dual cores, and a built in DDR memory controller, fixing what could be all the flaws of the original G4. Why switch? Did Steve piss off Motorola too?

Like texanpenguin says, the PowerPC is really starting to take off. With all the next generation consoles using the architecture, you can bet that things will only get better. PowerPC chips could come down in price due to the extra volume. Compilers will definitely get better as the demand increases for better compilers than the current GCC. In short, this is the worst time to drop PowerPC.

Sometimes, I think Steve has 'the Voice'. It's amazing how he is able to convert the thinking of the masses in less than two days. The G5s are good chips.

[Cat]

Attention Rumour crossing!

from Macbidouille:

IBM has been facing huge problems with the PPC975, lately renamed 976 then 980 that has been continuously postponed.
The 970MP processor project has been launch to avoid a big gap and a abnormal delay between processor update. But it would have been ready only by beginning of 2006 at the earliest, while IBM is still facing problems to manufacture the PPC970FX in large volume.
Concerning the PowerBookG5, all developments have been stopped. The reason: no way to manufacture PPC970GX necessary to answer both requirements for mobility and battery lifespan. In addition, they could not setup an efficient cooling system as well as a corresponding battery.
IBM has also decided to drop the PPC750VX which was designed to replace the G4 in entry-level computers.

Like Steve said: no 3GHz, even after two years, no hope at all for G5 PB.

On the bright side, read:

http://arstechnica.com/columns/mac/mac-20050608.ars

What about a Dual Dual-core 64bit intelliMac @ ~4GHz? Doesn't sound so bad ...

[Macaholic G5]

To quote H. G. Wells;

Adapt or perish, now as ever, is nature's inexorable imperative.

[Viro]

Originally Posted by Macaholic G5

To quote H. G. Wells;

Will _all_ adaptation guarantee survivability?

[chornbe]

Originally Posted by Ceroc Addict

What bummed me out was that this news took up almost the complete keynote (made up for by the aside that it'll be possible to add labelled chapter markers to podcasts). I wanted some kind of product announcement (e.g. as long as they're going Intel, they might as well have announced that they were making a four button mouse too ).

ah, but this was a developer conference, not the typical marketing/launch conference, tho'.

[fjdouse]

Originally Posted by Viro

Will _all_ adaptation guarantee survivability?

Nice (also love the proverb in your sig)

I agree with your previous observation that this was a really silly time to do this.

So why do it?
1. Steve is RIGHT about the future?
2. Steve's ego was bruised and fought back?
3. Big business/money is at work and we have yet to see how it will all turn out.


I know a lot of users, especially those who are non-technical don't care about this, nor should they. I still feel that if done wrong, this move will weaken rather than strengthen the Mac.. or Apple PC....

[metro10]

So I re-read page 459 in the old Big Book.... Life on life's terms. I just really liked the Power PC 64 bit chip and all it promised. However, this is almost as bad as when Quaaludes were outlawed Probably just as well and yet one more thing to get over!

[fryke]

I think my position was different even _before_ Steve showed up on that stage, and that probably helped me a lot to get over it. Living and breathing in the Mac rumour world for a long time, I knew that it was technically more than possible to do such a step. And I sat there and listened, and I said to myself: "If Steve can tell me it's _not_ going to be a harder transition than the ones before, then I'm alright with it." And from this position, Steve did a brilliant job, really. He not only told the devs _that_ it would happen, he also very clearly showed them the way. Go Xcode. Create Universal Binaries. It was a clean, smooth job done well. However: It was also clear that while developers would probably see this as a challenge and opportunity, consumers (not all of them, of course) would see it as a challenge and danger.

I guess now that he's done the right job with the devs, there'll come a show or two where Steve has to to de good job with the customers/consumers.