Macintosh Moving to Intel Processors [Updated]

[steeldrivingjon]

This is at the root of it all.

As for OSX always been compiled for more than one chip....this is a multi-squillion dollar portable music player company we are talking about. They can afford to dedicate a lab to do this stuff, it's absolutely nothing to them.

Realistically, they had to port the OS from Intel to PowerPC, back after the merger with NeXT, before they could "port" it from PowerPC to Intel.

After all, it's based on what Apple bought from NeXT, and NeXT didn't have an up-to-date PowerPC version of their operating system. (They had a PowerPC version of a five or six year-old version of their operating system, which they built in the early 90s for a prototype of some hardware NeXT never shipped. NeXT's most current OS would have been running on Intel.)

 

[abluesky]

This is at the root of it all.

People are going to be for and against "the move", but it's not like Apple had a choice. So you can argue for and against all day. The fact is IBM shut the door and Apple got dropped into the Intel bucket.

I sensed a reserved tension in Jobs' speech. Digging up one ancient photo of Job's and Mr Intel-Whoever to show some sort of history. He could have done the same with Bill Gates. Hesitantly showing that, he has indeed, been running on an Intel processor all morning. And it is because of this that Photoshop took 3 mins 30 seconds to load through 'Rosetta'. And the story of calling some random software developer up in the middle of the night to fly him over just added to the 'hacked together last minute' sense the Keynote had for me.

As for OSX always been compiled for more than one chip....this is a multi-squillion dollar portable music player company we are talking about. They can afford to dedicate a lab to do this stuff, it's absolutely nothing to them.

The Intel move is a fact, there was no choice in the matter, and it sucks.

I just watched the video of the keynote again, and to me it looks like it took about 30 seconds for Photoshop to load through Rosetta, which I think is pretty damned impressive. It certainly didn't take 3:30...
...

 

[Ravenflight]

Apple has a miniscule market share, and it is almost entirely limited to consumers and hobbyists. Even though they have a corporate presence in the Washington DC area (a building in Reston near Dulles Airport sports the Apple logo), they have no detectable presence at all in the very lucrative, high-volume, price-sensitive government market. I wouldn't be surprised to find out that there are more Windows machines in the Navy-USMC alone than there are Macintoshes in the whole world. Has any ever heard of a TEMPEST version of the Macintosh?

Apple doesn't have the market share, the pricing, the product line, or the corporate relations programs that are necessary for it to even think about "kicking Microsoft's butt." Fortunately, they don't have to.

On the day that the Apple web site asks you to choose whether you are a home user or a business user, you'll know that Apple is on the way up.

I was unaware that NASA/JPL was no longer affiliated with the US Goverment, or was not considered a "Goverment Market".

My point was that while Apple's computer sales have grown by 40%, they will really take off after this. Suddenly when people buy a Mac they no longer have to choose to give up Windows. They can say "well if I buy an Apple I can run that OSX that every PC magazine says is the greatest thing since sliced bread, and I can still install Windows on it later if I need to" Except they never will. Because once they buy the Mac x86 they'll ask themselves what they need the 200.00 Windows XP Home Edition for, probably the answer will be Microsoft Office or some other Microsoft program that already runs on OSX- so what's the point of buying a second inferior buggy OS? Dell & Co will also be in serious trouble, because all of a sudden they have a new competitor making x86 PC's that are more stylish, integrated, and problem free than theirs will ever be.

 

[Mord]

ASUS, Gigabyte sound better to me

VIA makes chipsets not motherboard, there are three main chipset brands ULI, VIA and nividia (ATI is makeing some now) nvidia are the best, DFI make the best boards for overclocking (gigabyte, EPoX and chaintec are fairly good too).

 

[Mord]

VIA does not suck and even Nvidia uses VIA chips in Nforce MB's.


Hey Hector where u been u missed alot these past couple of days. I've been wanting to gload about the demise of your precious PowerPC.

P.S. I want my Centrino Powerbook.....NOW

i was talking from my and my friends experience, also i have not missed allot these past few days i have probably posted more than you have it's just been swamped out, anyhoo, intel must have something pretty freaking amazing out in 2006/7 for apple to switch to them, or IBM's progress with the 970GX/MP must be really slow, apple is not switching for intels current netburst chips they are switching for what intel will have in 2006/7 which must have looks a heck of allot sweeter to apple than whatever IBM had going for them.

personally after getting over the loss of the ppc this has created the perfect computer for me, one that i can boot Mac OS X for work windows for games and linux for geeky work.

 

[Counter]

I just watched the video of the keynote again, and to me it looks like it took about 30 seconds for Photoshop to load through Rosetta, which I think is pretty damned impressive. It certainly didn't take 3:30...
...

I exaggerated slightly.

My G4/400 loads Photoshop CS as fast as Job's Mactel did in the Keynote.

I guess he must have had more fonts and plugins...

 

[abluesky]

I exaggerated slightly.

My G4/400 loads Photoshop CS as fast as Job's Mactel did in the Keynote.

I guess he must have had more fonts and plugins...

Slightly? I would say 3:30 as compared to :30 is more than a slight exaggeration. I think this just plays into certain peoples misconceptions.

I was actually very impressed with the whole keynote, including the performance of the system he was using.

The sky is not falling.

 

[Fabio]

Why not make this transition permanent? (rather oxymoronic I know)
Forget ending the transition, just make Macs that can use either PPC or x86.
Before you say I'm crazy, think about it: as Steve is saying:
1) OSX is fairly cpu agnostic
2) Shops are going to have to live with a double fat universal binary for quite some time.
3) The translator software is put in place, are they going to yank it? Why not expand it to go the other way too?
4) Apple can pick the best CPU for the task at hand, just like any other part.
5) I don't think anyone would be confused, or grieving, or sad, or in shock... but would be the evolution of computers from the revolution company. Steve would be again hailed as a visionary, instead of, "I guess he had to face up to reality"... come on, this is STEVE JOBS we are talking about here. IMPOSSIBLE == POSSIBLE, right?

Now, as is, IBM has ZERO incentive to improve the 970 for Apple.
But there's the what if, a BIG WHAT IF, something cool and new comes out from IBM, or Freescale after the transition is complete? Oops, hey group, let's go back?
IBM made the 970, and it is really good. Why can't they do it again?

A "transition" will scare people into holding out, hurting current sales no matter how many people say it doesn't matter. If it's not a transition, but spun as more of it doesn't matter anymore, then maybe people will think, hey cool, it's just like picking different video cards or something. Mac is great at hiding the complex. Most people in this thread have shown they don't want to worry or even know the details in the machine, just that it works, and it works well.

Nothing new, or different is done in this transition, this isn't Apple's first time at it either. But, making the Mac support 2 different cpu architectures at the same time is really cool, and would clearly put OSX into a category untouchable from that other one.

Downside is the support required, but over time this would just be the standard, and a massive competitive advantage.

Steve, if you can hear me, never end the transition! Long live the transition!

-Wyrm

You hit the point! It seems very clear to me. Steve didn't say (I think it's not a case) that they will quit with PPC Macs, and that's the smartest thing to do, to keep the door open and being able to choose the best architecture for each product. Of course they need all the developers to start porting their software to a "universal" binary format. Steve couldn't say this because to be successful they need the developers to start massively the port. As ever, he didn't lie, he just didn't say all the truth!

 

[Counter]

Slightly? I would say 3:30 as compared to :30 is more than a slight exaggeration. I think this just plays into certain peoples misconceptions.

I was actually very impressed with the whole keynote, including the performance of the system he was using.

The sky is not falling.

If I had exaggerated any less people may have thought it wasn't an exaggeration.

This is enormous and for me Jobs didn't touch on anything enough. Jobs for me came across like somebody told him "were on intel" this morning. "Intel, Intel.....uhhh.....*flicks through photo album*....right, that's going in. Erm, benchmarks.....*couple of honchos say 'erm leave them out' at the same time* Roadmaps! IBM's can be shown as a closed door, Intels an open one. 'Make it sound better than that'. How can I make yes & no into a benchmarky sort of graph where our stick goes way up there. Performance per watt or something.

 

[abluesky]

If I had exaggerated any less people may have thought it wasn't an exaggeration.

This is enormous and for me Jobs didn't touch on anything enough. Jobs for me came across like somebody told him "were on intel" this morning. "Intel, Intel.....uhhh.....*flicks through photo album*....right, that's going in. Erm, benchmarks.....*couple of honchos say 'erm leave them out' at the same time* Roadmaps! IBM's can be shown as a closed door, Intels an open one. 'Make it sound better than that'. How can I make yes & no into a benchmarky sort of graph where our stick goes way up there. Performance per watt or something.

In a year or two the numbers will speak for themselves. I don't think you will have anything to worry about with regard to raw performance. SSE3 (Intel's version of Altivec), EM640, dual core, quad core, quad core MP, you will see huge leaps in performance over todays Macs or Intel based machines.

But, then again, if he had shown you a pretty chart that said "our machines are going to be X times faster". You would have said, why should buy the current generation of Macs then?

In the end, this was for developers not for end user consumption and I think for that audience, it made sense and was a good keynote.

If you look at the roadmap and you consider where Apple wants to go, I think they made the correct decision - but we will all find out over the next few years...

Cheers!

 

[Quartz Extreme]

Jobs for me came across like somebody told him "were on intel" this morning. "Intel, Intel.....uhhh.....*flicks through photo album*....right, that's going in.

[Sarcasm]Thats why they've been compiling Mac OS X for Intel for 5 years.[/Sarcasm]

How can I make yes & no into a benchmarky sort of graph where our stick goes way up there. Performance per watt or something.

They made this decision based on the fact that IBM hasn't been able to produce a version of the G5 that uses low enough power to run in a laptop, hence, performance per watt, it's not that difficult to understand.

 

[woolfgang]

Irrelevant. At the time of those benchmarks there were hardly any apps that took advantage of 64-bit.

It is hardly impressive that a 64-bit cpu running 32 bit apps can outdo a 32-bit cpu running the same.

It is PURELY a marketing ploy.

You just contradicted yourself. Read very slowly next time.

 

[Counter]

If you look at the roadmap and you consider where Apple wants to go, I think they made the correct decision

Apple had no choice.

[Sarcasm]Thats why they've been compiling Mac OS X for Intel for 5 years.[/Sarcasm]

[Sarcasm]They've probably been compiling for Xbox too. Just in case.[/Sarcasm]

I may not know all my tech, but I have used PPC since it first came out. I am now in the market for a new Mac, having used PPC machines for 10 years, and I'm not sure what to do. I want to know the machine I buy has 5 solid years of longevity (and a sale above ten bucks would be nice after) and feel I haven't been sold on this. And, I think I represent a large chunk of the Mac buying market.

 

[Ravenflight]

They made this decision based on the fact that IBM hasn't been able to produce a version of the G5 that uses low enough power to run in a laptop, hence, performance per watt, it's not that difficult to understand.

I totally agree that that was Apple's reason, but it still leaves open the question why they couldn't show where Intel's performance would be x amount better than PowerPC (even theoretically) by the time they start introducing these chips. My feeling is they will be x amount better than G4s (in laptops) but they aren't expected to be significantly faster than future G5's for at least 1 1/2 years- hence the reason they won't start with Power Macs and work down, they'll start with the lowest performing laptops/Mini and work up.

 

[GTKpower]

You just contradicted yourself. Read very slowly next time.

A 64-bit architecture chip that runs 32-bit apps very well

vs.

A 32-bit architecture chip that runs 32-bit apps not as well (aging P4 design.)

So, i should be impressed because . . . . . . . ?

 

[Earendil]

Apple had no choice.



[Sarcasm]They've probably been compiling for Xbox too. Just in case.[/Sarcasm]

I may not know all my tech, but I have used PPC since it first came out. I am now in the market for a new Mac, having used PPC machines for 10 years, and I'm not sure what to do. I want to know the machine I buy has 5 solid years of longevity (and a sale above ten bucks would be nice after) and feel I haven't been sold on this. And, I think I represent a large chunk of the Mac buying market.

A large chunk that doesn't include the majority of the developers at WWDC or as far as I can tell, this forum. And the greater Mac community at large doesn't have a friggin clue, couldn't tell a PPC chip from a x86 based chip, and doesn't need to!

For the last time, there are going to be FAR more PPC based Macs even two years after the transition. Apple nor developers aren't going to abandoned PPC. We already know that the next MacOS is going to be native to both chips. So calm down and buy a computer. You probably won't use it for more than 5 years anyway.

~Earendil

 

[Quartz Extreme]

I am now in the market for a new Mac, having used PPC machines for 10 years, and I'm not sure what to do. I want to know the machine I buy has 5 solid years of longevity (and a sale above ten bucks would be nice after) and feel I haven't been sold on this. And, I think I represent a large chunk of the Mac buying market.

Current PPC Macs will still have many more useful years ahead of them. As Jobs said, there are new PPC products still in the pipeline (970MP?). And, developers can compile universal binaries so their programs will run on the Macs of today and the Macs of tomarrow. I sympathize with your current purchasing dillema, but Apple would never forsake their current installed base.

Example: the newest Mac OS still worked on 68040 chips until 1999 with OS 8.6. That's five years after the first PPC Mac was introduced...

 

[GTKpower]

Current PPC Macs will still have many more useful years ahead of them. As Jobs said, there are new PPC products still in the pipeline (970MP?). And, developers can compile universal binaries so their programs will run on the Macs of today and the Macs of tomarrow. I sympathize with your current purchasing dillema, but Apple would never forsake their current installed base.

Example: the newest Mac OS still worked on 68040 chips until 1999 with OS 8.6. That's five years after the first PPC Mac was introduced...

It's no use, you and I can say this until we're blue in the face.

The core Mac community will STILL feel the sting of this transition, no matter how useful their computers will be in terms of GETTING WORK DONE. Even a low-end PC will get the job done in terms of e-mail, word-processing, internet, and even some games.

The point is that all the Mac-lovers will no longer feel "cool" and "special." They were under the impression that they were part of some fabled elite, some mystical group that had the inside track, whereas the rest of us were just idiots who were too cheap and unimaginative to buy into Apple's vision.

Meanhwile, us "idiots" are running high-end, quality 64-bit AMD rigs, with the latest ATI and Nvidia videocards, with oodles of RAM for a fraction of the price. And we can sawp out our cpus to extend the life of our systems. Unfortunately, the integration between hardware and software is not the same as Apple's. That's alright. In recent years that has become an ephemeral issue.

And then "Apple's vision" suddenly met "the real world." And it wasn't a very beautiful friendship.

Until Apple puts up their sales figures for 2006/2007, they have no claim on anything, particularly the patently unfounded claim (in 2005) that they'll bury Longhorn. If an expensive, closed archtecture didn't do it from 2001-2005, it won't do it from 2006 and beyond. Especially not when Linux has shown what you can do with a server, and what you can do with little money and average hardware.

Unless of course Apple can get OS X to run on an ipod with a detachable keyboard.

Now THAT, would make for quite a keynote.

 

[Quartz Extreme]

The point is that all the Mac-lovers will no longer feel "cool" and "special."

It would take more than switching processor architectures to do that. Apple is so much more than IBM PowerPC. The design of the software, the operating system, the computers themselves....that is what Apple is about.

IBM, Freescale, Intel, AMD....whatever.
A Mac will always be a Mac.

 

[HelloKitty]

In theory, Apple could put PowerPCs in some Macs, and Intels in the others. If IBM made an improvement to the G5, and got the speeds up and temperatures down, Apple could put them in desktop models and rack models geared for graphics or science. The laptops and other models could keep using Intel.

Actually, come to think of it, Apple *will* be putting PowerPCs in some Macs, and Intel in the others, until they finish the transition around the end of 2007.

I just mean that Apple could do so indefinitely, choosing the best processor for a given role at a given time. Whichever processor is best for a laptop, or a graphics workstation, or a server, at a given price point, at a given time, could be used, without much concern over whether it is a PowerPC or an Intel chip.



When Steve was running NeXT, they used to support *four* very different CPU types, and three of those were in hardware that NeXT didn't control. And that was a tiny company compared to Apple's resources.

That's what I hope could happen. I hope Apple can still keep PowerPC as an option and not to shift to Intel completely. Because I think PowerPC and (or IBM) is still very capable of making fast CPUs. Personally, I have no problem using a PowerPC Macintosh at all.

Now I feel very guilty complaining about the minor 0.2Ghz update a few months ago. If you're watching, Steve, I really didn't mean that, we'll tolerate minor updates rather than switching to Intel..Please don't do this to us..

 

[jZilla]

What I'd love to see as soon as possible...

An in depth Jobs article where he lets us have some more of the vision.

I don't expect it by the way, I's just like it. Oh, and get Ive in on it as well.

Ta!

 

[BOOMBA]

Now I feel very guilty complaining about the minor 0.2Ghz update a few months ago. If you're watching, Steve, I really didn't mean that, we'll tolerate minor updates rather than switching to Intel..Please don't do this to us..

This is the stupidest thing anyone has ever written.

What is the beef against Intel?
Their processors ARE faster than the PPC, and in 1 year when they switch Apple will have upgraded boards and buses and made OS tweaks so the machines are going to be much faster than what would have been had they not switched.

You were right to complain about very minor sped bumps over a TWO YEAR period.

Why would you want to stay with slower machines simply because in your mind Intel = Microsoft?

(And by the way, Microsoft isn't evil either.)

 

[chibianh]

This is the stupidest thing anyone has ever written.

What is the beef against Intel?
Their processors ARE faster than the PPC, and in 1 year when they switch Apple will have upgraded boards and buses and made OS tweaks so the machines are going to be much faster than what would have been had they not switched.

You were right to complain about very minor sped bumps over a TWO YEAR period.

Why would you want to stay with slower machines simply because in your mind Intel = Microsoft?

(And by the way, Microsoft isn't evil either.)

yes.. what IS the beef against Intel? I still haven't gotten a clear answer to that question.

 

[BOOMBA]

the REAL question

To me the real question is going to be COST.

(Unlike some people I have already accepted these machines will be faster and this will be a smart decision).

I have read "experts" commenting that the move to Intel will likely not drop the price of a Mac.

My 2 questions here are:

1) does a dual core chip cost the same to manufacture as 2 single core chips?
2) does the Intel chipset support more than one dual core chip running?

in my mind, these chips have got to cost almost half what the PPC chips do.
(this is based on no real information, just my imagination).

So if this is the case, I either want $500 off the price of the machine (another figure I pulled from my arse) or I want TWO dual core chips in my PowerMac.

Thanks, you've been great.

 

[Quartz Extreme]

does the Intel chipset support more than one dual core chip running?

I believe they have something to this effect on schedule for 2007.

They also have a quad-core, 64-bit chip intended for high-end desktop / server computers scheduled for late '07. (I'm thinkin' Xserve and PowerMacs...)
http://www.xbitlabs.com/news/cpu/display/20050505142821.html