The Big Mistake
First let me preface by saying that His Steveness is an awesome salesman. No doubts, no questions - simply awesome.
However, IMHO he did make a mistake with the Intel announcement on Monday.
What faux pas did he commit? He made a blunder that usually an engineer would make, not a sales/marketing person. He told the truth
By telling the community outright that the PPC was going to be replaced by Intel over this transition period, he turned the PPC platform into a Lame Duck. Some will call it The Osborne Effect as well but the end result is the same: FUD within the community as a whole, and a potential unwillingness by the buying public to continue purchasing PPC-based machines in the interim.
If he had been thinking clearly, again IMHO, he would have "spun" it totally differently. How? Allow me to paint the picture I feel he should have shown ...
"Today I would like to announce that what some of you may have read in the WSJ and on CNET is true - sorta. Today I would like to announce that the Macintosh is going to begin reaching out to more consumers, more businesses, and more markets. How will we do this?
By being FLEXIBLE.
If you were to come to me and say 'Steve, I'm out of shape and inflexible - what would you recommend?', I would tell you to take a yoga class - that would help your flexibility don't you think? <waits for laughs>
Well today I'll be asking you, our developers, to help Apple be flexible and go with us to that yoga class.
Today I would like to announce that we are adding some flexibility to our products by introducing Intel microprocessors to our hardware platform. <pause for dramatic effect>
After looking at both the PPC and the Intel roadmaps for CPU development, it became clear that we needed to tap the strengths of both. PPC is an awesome processor for our PowerMac and XServe platforms. In creative enviromnents and in advanced computational situations, the PPC is an incredible price-per-computational unit value.
But in the portable and consumer space, the long term vision of the PPC just didn't seem to get where we wanted to go. We can envision a number of products we would like to make for you in the future but we just can't see how to do it on the PPC roadmap alone.
Enter Intel. With a processor roadmap that includes low-power, low-heat products that perform at dizzying levels in the portable and consumer products, they make a great fit for our PowerBook users, eMac, and even iMac users.
<gives the transition speech about how they've done this before>
Now keep in mind, we are not leaving the PPC. It will still be in place in our PowerMac and Xserve products and we will continue to bring some amazing PPC based products for you in the future. But for our needs, for our customers needs, the Intel processors are a better value in the portable and consumer space.
It sounds difficult I know. Many of you are sitting there shocked and stunned - and not a little amazed I'm sure. 'How will we do this Steve? Have you lost your mind? We just did a transition from OS9 to OSX, now you want us to do Intel as well?'
A daunting task at first, until you know what I know. And what I will now share with you..
<gives speech about OSX on Intel for 5 years, etc>
<tells them about Xcode 2.1 and how you just click the button for Intel compile, etc>
So now that I've dropped this on you, let me drop One More Thing - all the demo's you've seen so far? All the slides you've seen so far?
All have been running on an Intel based Mac.
<shows About screen, etc> <waits for gasps, then applause>
<talks about Universal Binaries, etc>
And now to further ease your mind, allow me to introduce you to one of our developers who has already done this ... <introduces Theo Grey of Wolfram>
<keynote goes on from here>"
So, IMHO, he should have spun it as Apple branching out by adding Intel in an attempt to broaden the appeal and hardware flexibility of the platform. Increase marketshare, lower Mac prices, etc. ANYTHING but tell them that PPC was going away. Then, after 2 years or so, quietly phase PPC out of the product line entirely. KEEP developers making Universal Binaries as a matter of course. Don't let them compile/develop for just just one, that way, IF you ever wanted to use a future PPC chip, it would already be a done deal.
Just my $0.02
BnB
However, IMHO he did make a mistake with the Intel announcement on Monday.
What faux pas did he commit? He made a blunder that usually an engineer would make, not a sales/marketing person. He told the truth
By telling the community outright that the PPC was going to be replaced by Intel over this transition period, he turned the PPC platform into a Lame Duck. Some will call it The Osborne Effect as well but the end result is the same: FUD within the community as a whole, and a potential unwillingness by the buying public to continue purchasing PPC-based machines in the interim.
If he had been thinking clearly, again IMHO, he would have "spun" it totally differently. How? Allow me to paint the picture I feel he should have shown ...
"Today I would like to announce that what some of you may have read in the WSJ and on CNET is true - sorta. Today I would like to announce that the Macintosh is going to begin reaching out to more consumers, more businesses, and more markets. How will we do this?
By being FLEXIBLE.
If you were to come to me and say 'Steve, I'm out of shape and inflexible - what would you recommend?', I would tell you to take a yoga class - that would help your flexibility don't you think? <waits for laughs>
Well today I'll be asking you, our developers, to help Apple be flexible and go with us to that yoga class.
Today I would like to announce that we are adding some flexibility to our products by introducing Intel microprocessors to our hardware platform. <pause for dramatic effect>
After looking at both the PPC and the Intel roadmaps for CPU development, it became clear that we needed to tap the strengths of both. PPC is an awesome processor for our PowerMac and XServe platforms. In creative enviromnents and in advanced computational situations, the PPC is an incredible price-per-computational unit value.
But in the portable and consumer space, the long term vision of the PPC just didn't seem to get where we wanted to go. We can envision a number of products we would like to make for you in the future but we just can't see how to do it on the PPC roadmap alone.
Enter Intel. With a processor roadmap that includes low-power, low-heat products that perform at dizzying levels in the portable and consumer products, they make a great fit for our PowerBook users, eMac, and even iMac users.
<gives the transition speech about how they've done this before>
Now keep in mind, we are not leaving the PPC. It will still be in place in our PowerMac and Xserve products and we will continue to bring some amazing PPC based products for you in the future. But for our needs, for our customers needs, the Intel processors are a better value in the portable and consumer space.
It sounds difficult I know. Many of you are sitting there shocked and stunned - and not a little amazed I'm sure. 'How will we do this Steve? Have you lost your mind? We just did a transition from OS9 to OSX, now you want us to do Intel as well?'
A daunting task at first, until you know what I know. And what I will now share with you..
<gives speech about OSX on Intel for 5 years, etc>
<tells them about Xcode 2.1 and how you just click the button for Intel compile, etc>
So now that I've dropped this on you, let me drop One More Thing - all the demo's you've seen so far? All the slides you've seen so far?
All have been running on an Intel based Mac.
<shows About screen, etc> <waits for gasps, then applause>
<talks about Universal Binaries, etc>
And now to further ease your mind, allow me to introduce you to one of our developers who has already done this ... <introduces Theo Grey of Wolfram>
<keynote goes on from here>"
So, IMHO, he should have spun it as Apple branching out by adding Intel in an attempt to broaden the appeal and hardware flexibility of the platform. Increase marketshare, lower Mac prices, etc. ANYTHING but tell them that PPC was going away. Then, after 2 years or so, quietly phase PPC out of the product line entirely. KEEP developers making Universal Binaries as a matter of course. Don't let them compile/develop for just just one, that way, IF you ever wanted to use a future PPC chip, it would already be a done deal.
Just my $0.02
BnB
Comments
Ie, it'd be Carbon all over again. We've been down the route. Apple put tremendous amounts of energy into bending over backwards for the Carbon folks, and Monday the bill was handed to them. Time to move forward, get with the program, and move ahead.
Once the apps are Universal Binary ready, then *any* platform that Apple wishes to target is a possibility. Long-term flexibility is a must, as you point out, but most people are too short-sighted to see why they should have to put any effort into it. This forces them to. I think it was the right move.
Originally posted by Banana Nut Bread
Then, after 2 years or so, quietly phase PPC out of the product line entirely.
Just my $0.02
You're being generous.
Originally posted by Banana Nut Bread
What faux pas did he commit? He made a blunder that usually an engineer would make, not a sales/marketing person. He told the truth
I do however feel that Steve didn't talk up the PowerPC enough to comfort today's buyers. I feel he should have said something along the lines of "Our PowerPC products of today and the next 24 months are great; it's the years after that we're looking out for". He eluded to that with Roadmap talk, but didn't quite say it that bluntly.
Honestly, PPC support is going to quickly fall by the wayside. This announcement was essentially a sword through Altivec's heart. Installed base or not, the future is out there.
Originally posted by groverat
The plan is to move all of Apple's eggs from one basket to the other, not to split them up amongst two baskets. Sure, we'll all know in the back of our minds that we could go PowerPC if necessary, but it would be great to see Apple maintaining a positive relationship with IBM & Freescale. Honestly, what could it possibly hurt?
Honestly, PPC support is going to quickly fall by the wayside. This announcement was essentially a sword through Altivec's heart. Installed base or not, the future is out there.
Yes it could resume in :
universal binaries, and stop developping for Altivec
The framework offers an API for the most common vector calculations in mathematics, graphics, sound, DSP use, etc. When compiled for PPC, it produces AltiVec instructions optimized for that particular chip. What, you thought all PowerPCs had the same AltiVec implementation? Au contrare. Just as with the standard PPC instructions and instruction ordering details, different models vary slightly, resulting in different optimal optimizations. Accelerator took care of that for the dev.
Developers that used Accelerator are now pretty much set - Apple converts the framework to emit SSE/2/3 depending on the final chip being targeted, and the dev does damned little.
Those that didn't, might want to look at it again.
Those who simply *couldn't* due to the needs of their app, or the lack of support in the framework for what they do, are going to have some serious work to do... but I have to believe that's the minority of cases.
Originally posted by Kickaha
Au contrare.
Au contraire
Originally posted by snoopy
Au contraire
Does it really matter?