WWDC 2005 Keynote Notes

Show notes: Steve Jobs’ introduction of OS X on Intel processors, Mac growth, etc. Took forever to get a WiFi connection. 3800 attendees, 3700 of whom have PowerBooks, so not much bandwidth to go around. This will be old news by the time this goes up, but FWIW…

Non-Intel-related cheerleading first:

Podcasting will be built into iTunes+iPod – direct subscriptions
Ability to subscribe to New Music Tuesdays
Music artwork changes as you scrub through – jump through feed via thumbnails

Growth of PC market has dropped over past few years, Mac growth 40% in the same period.
Mac growth is 3x the rate of the rest of the industry.
2,000,000 copies of Tiger already shipped (6 weeks)
Leopard is next release.

Widget: Rabbit Radio finds local NPR station; Yahoo’s traffic updates. 400 Widgets available already.

QT7 for Windows available today.

Transitions: Summary of transitions from 68k to PowerPC, and of OS9 to OS X. This one is going to be easier than those were.

Apple can’t deliver 3GHz PowerMac or G5 laptop as promised/hoped. Apple can’t build the software they want on PPC because the performance of the hardware is not keeping step with goals.

Performance per watt: how many units of performance per watt? Intel gives 70, PPC 15.

June ’06 – should be shipping Macs with Intel procs.
June ’07 – transition should be complete.

Two challenges:

1) Make OS X sing on Intel. OS X has been leading a secret double life. All software designs since the launch of OS X have been processor independent. Every project, even the first release of OS X, was built dual-architecture. Everything Apple has done in software for the past five years has been compiled for both PowerPC and Intel “Just in case.” Wow.

Steve has been demoing on a Pentium system all morning.

2) “Your apps.”
Cocoa apps – a few minor tweaks and a recompile
Carbon – a few more tweaks and a recompile and it just works
Metrowerks — You gotta get to XCode.

XCode 2.1 ships today and includes PPC/Intel target build selector, and lets you build a “Universal Binary.”

PPC will be supported for a very long time – huge installed base.
Recompiling is a lot easier than carbonization.

Mathematica recompile: Mathematica includes the world’s largest collection of algorithms in all kinds of programming styles. “Code that hasn’t changed since the Reagan Administration… some of which I may have written.” They checked the Intel checkbox and had a copy of Mathematica running on Intel in two hours. The folks at Wolfram Research thanked Apple for freeing up their weekend.

Because not all apps will be recompiled in time, “Rosetta” will do dynamic binary translation. Rosetta is transparent to users – not like running Classic. Just click and run – users won’t even know. Demo: Word launched and ran without recompile. Perfectly fast. Same with Photoshop (and all of its plugins).

Develper transition kit – PowerMac w 3.6GHz Pentium 4, OS 10.4.1, XCode 2.1, priced at $999, must be returned to Apple in 2006. Available only to top-tier ADC members.

1996 – Apple set fire to Intel’s bunny man. (showed commercial). No love lost. In 1996, PowerPC kicked booty on x86 performance. It just isn’t true anymore.

Much better write-up of the show at Anandtech.

13 Replies to “WWDC 2005 Keynote Notes”

  1. So MacOS is migrating to Intel.

    I used to be a PowerPC biggot. But recently, as I’ve watched Powerbooks lag in performance to PC notebooks, and can’t see why IBM would be motivated to provide R&D for a few million PowerPCs per year to Apple when they’ve got Microsoft X-Boxes about to be sold by the tens of millions with the new cell processor inside. I’m sure one reason we don’t have a mobile G5 processor after 2 years is at least partly due to resouces spent on the big money-maker cell rather the G5.

    I’m very glad Apple is making this move. I would love to have a Centrino-powered Powerbook (now please!). Imagine how much faster Windows emulation will be on such a machine (VMWare or VirtualPC), when it’s absolutely needed.

    Two years ago I was at WWDC when the G5 was launched. I was psyched for that chip and glad Apple was using IBM’s technology. But IBM has not come through in either raw speed (I was there when an IBM rep promised 3 gigaherz by Summer, 2004) or mobile processing. If the Intel announcement had come a year ago, I would have thought it a mistake, but now it’s clearly a smart move.

    Hurry up with those new Powerbooks! ;-)

  2. I was one of the people who always said this would never happen. Now that it has, it seems obvious. At one time, a million PCs shipped per quarter was a huge number. That was 15 or 20 years ago and the prices of PCs back in those days reflected the development costs of only churning out a million PCs. The reality today is that PCs (in sum total) probably approach 50 to 100 million shipped. This has dramatically lowered their costs for everyone… that is except Apple. Apple is different because they use PPC. The embedded space (from Motorola) and the server space (from IBM) bought them a few years but neither of those companies was going to invest what Intel does to improve their platform specifically for PCs (Apple). Apple, to their credit, has probably realized this now for many years – thus the fact that OS X has always (in house) also run on intel hardware. If they were going to transition, they needed this to work and also to be developing a method of compiling for both archs during the transition period which apparently they have also done (and additionally developed Rosetta which will run old ppc apps on the intel arch). Apple needs to be associated with a company who is developing chips for their market and can do so cheaply because of volume and now they are. Man does that seem obvious in hindsight. My only worry is that we will be stuck with the limitations of x86 forever and also there is no independent arch for which to compete with Intel. I guess you could argue that x86 keeps churning on, at least for now, and perhaps AMD keeps Intel on its toes in terms of competition. Probably the most recent rounds of negotiations with IBM made it clear that they were not willing to invest that much into the G5 and that they would not be as competitive with Intel for pricing due to low volume. Further, the disappointments with G5 clock speeds and the problems in the portable space made Apple decide now was the time to do this. Let’s hope the transition goes smoothly for everyone.

  3. If Apple can:

    * Make its Macs Windows certified and keep the price to 10% above a Dell of the same processor
    * Prevent normal PC’s from running OS X,
    * Do a “try it for 30 days free” offer for Leopard with a partition-in-a-file option like VMWare, or a smooth dual boot option

    —>They could CLEAN UP while still maintaining their hardware sales.

  4. Mahmood, there is definitely no plan to let OS X run on beige (or black) boxes. Whether someone figures out how to run Windows on the new Macs remains to be seen — remember there will be custom ROMs etc. on the motherboard. But it’s not inconceivable.

  5. I doubt that Apple’s hardware prices will drop. Look at the high prices that AlienWare charges for their rigs.

    And I suspect that enterprising hackers will quickly have OS X running on their x86 machines.

  6. I just watched the the streaming QuickTime version of the keynote after reading news about it all day online. I think Jobs made a sound case for making this transition, and Apple seems to have done what is necessary to make it as easy as possible, having come up with the new version of XCode, Rosetta, universal binaries to take care of current PPC-based Mac users and future Intel-based Mac users, etc. after years of doing their homework. The statements from the folks from Wolfram, Microsoft, and Adobe were reassuring. And unless I’m mistaken, there was no mention of IBM or Motorola during his talk, Apple’s former partners in developing the PowerPC.

  7. Can anyone speculate as to why Apple went with Intel processors instead of IBMs upcoming Cell processor? The Cell processor is supposed to be killer for multimedia purposes. That sounds like an ideal chip for Apple.

  8. Scot, what bothers me is that people seem to think intel=can run windows. I can imagine emulation would be much faster, but I don’t expect these will be the same as garage-built PC boxes. Thank goodness for that too. I don’t spend $2,700AUS on a Mac for it to be the same as a hunk’o’junk PC box.

  9. Kristan, there’s a whole lot of misinformation and speculation out there, and I think that means Apple has said too little, has not done a good job of explaining themselves to the public. A lot of these questions could have been addressed explicitly at the Keynote, but weren’t. It’s a shame, but there may be good reasons for that (tied into marketing messages, PR, not closing doors to the future, etc.)

    I’m not sure what the fear is about a Mac being similar to a generic beige box. The Mac is already full of network cards, sound chipsets, video chipsets etc from non-Apple sources. So the CPU is by Intel rather than IBM — who cares? Everything else remains *exactly the same*. That’s the message that seems not to have gotten through. Everything remains exactly the same. It seems there’s a lot less interest in this topic here at WWDC than there is “out there” on the Web. I think that’s because the attendees have seen so clearly that everything remains *exactly the same* and are not concerned in the least.

  10. > The Cell processor is supposed to be killer for multimedia purposes.

    Yes, especially when it would kill laptops with its heat. :)

    Problems were two:
    1. The Apple-version of the G5 (which is not the same as XBOX’s) was not receiving enough love from IBM to get progressed faster.
    2. The G5 and any other POWER-based technology were never released with laptops and power consumption in mind. Apple’s computer business ARE laptops primarily. And IBM just didn’t deliver in this department.

    So, Apple went with whoever who could. Intel in this case.

  11. Scot, it seems that a Mac differentiates from a ‘beige box’, is the design, and then the ability to run Mac (oh, and the price! Heh). If those are removed, what’s to stop you buying a $600 box, instead of a $2000 box to run Mac?

    Wouldn’t they lose their hardware profit margins?

    I think they’ve done very well, considering the options they had. I personally would love to see them tackle Windows, and run on OSX on my box here now, but probably won’t see that happen naturally, but with some odd-hack/patch. It’s good for some, bad for others, but in the long run, it appears to be good for Apple.

    Now they’ll have powerbooks worth buying, with speed margins greater than their ibook counterparts (rather than just more RAM, etc for $2000 more…).

    I think it’s a good, competitive move.

  12. > and then the ability to run Mac

    I presume that by this you mean “the ability to run OS X” ?

    So what you’re saying is, “Remove the ability to run OS X and the design and what’s to stop you from buying a $600 box.” If that’s what you’re saying, I don’t understand what you mean. Seems like you’re saying “Apple should keep selling Macs.” Well, um, yeah. If I’m misunderstanding you, care to clarify?

Leave a Reply

Your email address will not be published. Required fields are marked *