Apple confirms switch to Intel - Page 2

Is it, perhaps, because you have no technical reasons? Just like Steve Jobs gave no technical reasons or any insight into this supposedly "superior" roadmap that he claims Intel has lined up? I mean last year he was talking about the "superior" PowerPC architecture, but it's quite obvious he wasn't all that serious about its "superiority," so please do indulge us, what are the reasons that this should be considered a "great move"?

Ok, now that the major shock and awe of Steve's announcement has passed through me some thoughts come to mind.

Leopard in late 2006 (or early 2007):

So, from R. Kipling perhaps some hints:

http://www.sff.net/people/karawynn/justso/leopard.htp

I think that here Apple is drawing a metaphor for its own survival. Get back to where the game is. The apple (or leopard) must change it's spots to survive. Windows and Intel have kicked apple into a 3-16% market share. A small(ish) developer base. The leopard must change its spots.

Tiger is a hint of where apple is going and with this kind of great buzz about the OS there are plenty of switchers out there who have taken notice. they all have shitty windows boxes and tons of software that they paid hard earned $$ for. If the MacTel box can run those programs in some form of emulator then they are totally on board. The OS can hunt in the jungle again.

TIGER, tiger, burning bright\t
In the forests of the night,\t
What immortal hand or eye\t
Could frame thy fearful symmetry?

Fearful symmetry like running on PPC and x86?

Right now apple has a great platform for the high end. The dualies are fine for the purpose that they were designed. This should last us into 2007 when a wicked intel/powemac is debuted.

In the mean time the switcher platform will get an intel p4 chip in 2005. The mac mini = crack for the switcher base.

All of us kool aid drinkers will buy the apple hardware. Apple knows this. It will make it easy for developers to port their apps. If apple took its sweet time developing the tools for the developers to switch then watch what will become.

I am not afraid.

"What the caterpillar calls the end of the world, the master calls a butterfly." --- Richard Bach

I will stay with my G4 powermac until they give me a low power centrino with wimax. early 2006 will be it.

Unless Paris shows us some hardware.


+++++++++

Why Intel - and not AMD?

I wonder why the transition takes until mid-2007. Wouldn't it be wise on Apple's part to equip every possible device with the best price/performance processing options as soon as they can?

Maybe this has something to do with Intel's roadmap for high-performance CPUs.

There is one positive aspect to this alliance: If Apple uses the same processors Dell, HP and everyone else does update schedules should become more predictable.

But I do share many of the concerns mentioned in this thread:

- How will Apple be able to maintain its high margins when the industry average is far lower?
- How long will developers put effort into developing for both platforms.

This seems like a desperate move, but who knows how bad IBM hit the performance wall with the 970-series processors. One has to wonder why they did not consider using the Cell processor. But maybe we are in for a surprise and it will be used as a media coprocessor.

I am very excited as they times ahead should a few surprises.

I don't know what to think about this. Still shocked.

I really like PowerPC.

I imagine a great deal of this comes down to ego. Steve doesn't want to play second fiddle to another customer. I'm sure IBM diverted resources toward winning the game consoles, which led to less development on the Apple platform.

I can understand from IBM's standpoint with the consoles they will sell far more processor than with Apple. The question is was IBM ignoring Apple or just temporarily diverting attention.

To my observance IBM had the most innovative ideas and a clear vision of Power PC. 64bit - 32bit native processors were first from IBM, IBM first had the idea of dual core processors, IBM had the idea of virtualization.

To some degree I guess it doesn't matter if they have the idea and its not working in the Mac.

It just feels as though Intel is all over the place. Somewhat behind IBM and AMD on new ideas.

Intel chips are no cheaper than IBM's.

I'm not sure about this whole idea at all.

The wosrts possibility is of bootleg OSX running on beige Wintel boxes. I'm sure Apple is confident that won't happen. But hackers will try hard to break whatever they come up with. It's hard to see how they won't be successful in some way.

I don't know, I'm not convinced this is the right move to make.

Because it won't run Mac OS X.

First, Apple *could* keep G5+ PPC for high-end video, and transition consumer grade Macs to Intel. I know the downside...

Secondly, from a long-range strategic perspective, this is the transition that *must* occur in order for Apple to become the software shop it truly is. Apple, like many other manufacturers, have transitioned away from designing chip sets to sourcing them. What's left is the software. I've been saying it for years...it's the software stupid. The hardware is, erm, great because it's tightly controlled by Apple. They retain this very sort of control with today's announcement. What they gain is x86, access to DirectX 9 (imagine running an installer that asks for your Windows XP disk, and pulls out the DirectX components), and Intel's comparitively massive R&D and manufacturing muscle.

Bold. That's what this is. But I think relatively obvious in its necessity.

OK, help me out here:

--Jobs says the dual path engineering effort has been ongoing, the Mactel machines exist now.
--Jobs says porting Cocoa apps takes two hours, everything else, "a few weeks". Even allowing for PR happy talk, gotta assume the migration path isn't too hideous.
--Sounds like the emulation layer "Rosetta" exists now.

So, everything is in place, the trickiest part of this is managing the transition period, during which PPC mac sales are likely to tank. And.....

They will roll out the first Mactel box in a year? WTF?

I mean, I know everybody needs a chance to get their head around this, and the actual migration of apps will inevitably take longer than Jobs is making out, but if what Apple is claiming is even 50% true you would think they could start selling these things in a few months, not next June (and only a fraction of the line up by then, at that).

How many towers are they going to sell in the coming year? 50? Couple hundred? How many powerbooks? 3?

I really don't get it. Given the advanced state of the software effort (assuming that the whole dog and pony show wasn't utter horseshit), I really would have thought they would want to make the transition as fast possible.

Shit, if Rosetta works, at all, why not start selling the new stuff right now, along side the PowerPC legacy boxes, and let people choose to take whatever the emulation performance hit is, in exchange for knowing their purchase is in line with the future of the architecture and not a dead end?

Looks like those upticks in market share are going to be a fond memory by this time next year.

And what of sales of current hardware? Surely they will crash. Nobody wants to buy something which they know will radically changed/replaced in a year's time.

I know I was pretty damn close to buying an iMac G5. Steve just saved me a big pile of cash today.

Think different?

I don't trust all these quicky ports. I'n no programmer, but I remeber Adobe showing Photoshop that had been ported to OSX from OS9 and they did it over a coffe break. Yet how long did it take them to ship an actual product when the real switch happened?

I'm not saying this transistion can't be done, it is just that I've seen enough of this stuff not to beleive anything until it has happened.

I would like to know what hpppened to IBM. They fell from grace mighty quick. The best I can say is that if Intel doesn't deleiver then the Mac is in the same boat as the majority of the rest of the computing world.

Will be interesting to see how Mac hardware is priced with intel inside. It will be much easier to make apples to apples comparisons than the apples to oranges situation we have now.

Yep, my exact thoughts here. Who the fuck is going to by a PowerPC-based Mac even though Jobs said they will support both platforms for a while? I sure as hell am not. Geeks will find a cheap Sony or something just to run Rosetta/OSX86. Hell, it's probably on some P2P sites already.

(Anyone remember that Riddles of the Past... thread year or so ago? Rosetta was part of it, as I recall.

"The filling shall reveal a crust made from the Rosetta Stone."

Although faster, AMD's cpus are more expensive and AMD doesn't seem to have the volume that Intel has. Also I bet that Intel cut Apple a sweet deal for them to use Intel over AMD.

It is all about price. This is a good sign- Apple is going to probably be somewhat price competetive AND offer x86 CPUs.

Jeez, I must take up Apple prognostication professionally - Steve only missed a couple of points from my predicted script - must talk to him about that, he shoulda just followed my script like I told him to

I'm especially pleased and surprised to see they'll be shipping the development boxes - it will feel like 1993 again when I was porting NeXTStep 3.3 apps to NeXTStep/486, but this time with more hope of the company staying around!

I really think this is positive, especially with Rosetta - wonder if that's based on the Transitive stuff??? Anyway, I'll be very interested to see how apps like FCP run on it - if at all, or if I'll need to upgrade to get the Intel architecture.

Will also be interesting to see what they do about video cards. They should be able to use stock cards now (someone correct me if I'm wrong), I suspect that's true given the comments about people figuring out how to run Windoze on the Mac.

All I'm left to wonder at is what else is included in the deal with Intel - I'm hoping the deal is actually a platform deal a la Centrino and Intel is providing a whole chipset that supports PCIe, USB, FW, Bluetooth, WiFi, WiMAX (maybe) - esp is that chipset is also available for other manufacturers. That would really impact the price since Apple wouldn't be working with custom chips for North and South bridges

I wonder if I'll be able to run old Classic apps on an x86 Mac, with the x86 emulating a PPC emulating a 68040?

I think it about price, second. It is about availability and product outlook, first. That is what drove Apple from Motorola, first and IBM today.

I think we are all scared because of the unknown.

We will know more as the intel SDKs start to arrive on developers doors. Wonder if they will be resticted by NDAs to divulge any details of the OS performance. I hope not for Apples sake. They should try to quell as much fear and uncertainty as possible. Word of mouth from developers is the best way to do it.

I don't have much knowledge on how hacking works, but as far as vulnerabilities are concerned, does the chip on the machine play any part in it or is it all the OS? What about BIOS, must all intel architectures have BIOS? Isn't the BIOS another doorway for hackers, worms wtc into the machine?

Not really. OpenFirmware is to a Mac what BIOS is to a PC.

Their porting story for Developers is not that strong. It seems we need to shell out $999 for the oportunity to test on a PC running the OS, and - more than that - we cant compile or link on a ppc Mac.

The better version of this would be that, and....

a) Have the libraries for intel on Mac PPC so we can link against them.

b) Allow people to test on an Intel Mac at some local Apple store. or associated reseller.

Most Apple developers are small scale.

If Apple has OSX running on Pcs for 5 years then the clowns should market it. Didnt they learn a thing from Windows take over of the world without building hardware? Stop holding onto hardware so much and start selling what you do best. The Mac world is about the software folks its never been about the hardware and if you think so they suckered you. MR JOBS I WOULD LIKE 1 COPY OF MARKLAR FOR MY AURORA THANKYOU! 97% of the world is a lot of market.

This is really pretty good news.
1. Apple's notebook market is extremely important. The G4/G5 is going nowhere in this aspect. Apple had no choice but to switch to x86
2. It's going to be so compatible. The dual binaries are a great thing and neither PPC nor x86 will have problems. in fact, this means that Apple could even go back to PPC any time they want.
3. Intel will make Macs more compatible, cheaper, and have greater support. Imagine a NEW mac user. he goes into an apple store "so i like this mac mini, how much is it? okay, what kind of processor does it have?"

stop here! how is he going to react when the salesperson says "1.25Ghz G4" Okay, his last computer is a Pentium III at 1.4Ghz. He will think that he is buying a slower computer. But think about if the salesperson says "It has Intel Pentium M chip" He will think "Okay, this is a fast processor, I have one at work in my laptop for my PC"

then think about mac games. surely with x86, macs should be able to use standard PC graphics cards if Apple choses to do so. Even if that doesn't work out, think about running Virtual PC with no overhead. Even if nobody makes Mac games, you could run PC games on Virtual PC without slowdown. Cause, really, Mac will never get DirectX which is really what most games use.

Think about applications now. They can be converted from PPC mac to x86 mac in a matter of hours. You can get Unix applications and all that open source stuff. Not only will Virtual PC run so quickly, but surely anybody could port a Windows program to Mac in a snap. It's like having 3 machines in one!

Why Intel? Why not AMD?
1. Intel has a whole bunch of fabs. AMD only has one. Although AMD could meet the demand, AMD is really just wants to make chips and that's all. AMD can't give out the dirt cheap OEM deals that Intel has. The LAST thing Apple wants is to have so many supply/demand issues that they had with IBM.
2. AMD probably wouldn't support Apple's software plans. AMD wouldn't really want to collaborate, they'd just want to sell chips. Although I'd rather Apple use AMD chips, I like them better, AMD doesn't have the resources that Intel has.

I mean there's nothing really bad here for Apple. Even if Intel fails to deliver, Apple could switch back to PPC, or they could switch to AMD. They have put themselves in a really good position planning OS X around its cross platform abilities.

Apple's bottom line earnings tell otherwise.

I agree though, that OS X could generate a lot more income, but with such a small market share, they realize that hardware sales are needed.

I don't know much about hacking either but from one common vunarability are buffer overflows which means that any kind of code can be executed on the target system. Intel and AMD (VIA and others too, for that matter) try to cure this by declaring certain parts of the memory as "not executable".

Other security matters that IBM might have been reluctant to integrate include future DRM methods thus forcing Apple to change to be able to be granted the rights needed to expand their media strategy to movies (and beyond) by the media authorities.

So, if a hacker cannot access the powermac today, he/she will not be able to access the Mactel tomorrow? OR will she/he suddenly find that all the Wintel tools of the trade are now applicable to the Mactel? If that's the case, then Apple just took a huge leap backwards.

No I think OpenFirmware will stay around, basically as the system IO for booting, etc. Beyond that, however, is how those hacking tools will affect the Rosetta emulator. That is the big concern I have. I have dealt with fucking spyware/malware/viruses at work for the past 5 years and it is a tough battle. I don't want to have to dick with that crap at home too.

Download XCode 2.1. I think you'll find it contains all the x86 libraries and cross-compiles - From the release notes:


So, No, you won't have to buy the SDK, but if you want to test it thoroughly then Yes, you'll need it - from experience in the NeXTStep/Intel days, you'll find some odd bugs in your code that can only be found on an Intel box - almost always when you use C to manipulate bits and bytes directly. All the APIs will likely just work.

Is Apple going to use PC motherboards or are they going to still create Apple motherboards with Intel chips?

PowerPC could never go away...

IBM and Intel will be leapfrogging each other periodically - Apple should keep the fat binaries and both platforms around forever. When IBM is better, buy a PowerPC mac. When Intel is better, buy an x86 mac.

The keynote is playing here:

http://stream.apple.akadns.net/

Let's see... wait two years to buy a first gen AppTel PM or switch now to get used to the Intel product. Neither option makes me salivate, both to blow chunks.

And there are folks APPLAUDING this move???

Since Steve has been lying to us for the past five years, what guarantees do we have he hasn't also been paving the way for Mr. Gates to do an LBO (leveraged buyout) of Apple?

You know... to make the transition COMPLETE (heavy sarcasm).

I think it sucks that developers have to pay $999 for the transition kit!

APPLE this should be free!
You need as many developers behind your as possible and your pissing around with them.
Why should I partially re-write and recompile my code of OS X on x86. I can just as well re-write for Windows and appeal to a vastly larger market!

Fair enough from a hardware perspective. Apple has been dicked around by Moto and IBM regarding CPUs.
The MS vs OSX performance will be interesting in say 2009 when the next native x86 OS X release is out and more x86 tuning is done.

Dobby.

Don't charge develeopers for the transition kit!

Does this mean that currently 10MB applications will be bloated to 20MB or more? That would be very bad indeed, I think.

I feel sick about this. One of the advantages of having a different CPU was to pass the other Win/tel boys bye. Now were hard wired to the competition. MacOS X better run faster and better then longhorn on the same specked CPU or the Mac will end up like NeXT. Steve did you learn something from your last mistake? If any one can pull this off the Steve can, but is it possible at all?

I'm curious what the Pentium 4 processor Mac looks like...anyone have pictures?

What's the fuss about?

I've said for a long time, if I could buy tha PC with that spec, but install OSX - I'd do it.

Maybe now we'll see a little crossover - decent GFX, cheap RAM, PCI-X...

About time.

I certainly remember it, but I think that "Rosetta" was in reference to "Ink" or so the speculation was at the time...

But Geez, if that dude was right... way back then?

Wild to think of it, but I doubt it.

I am still amazed by this move... I thought Intel Processors were as big as firebricks!

--Hope Springs Eternal..... but my God....

Mandricard
AppleOutsider

They probably are "giving" it away (i.e. for cost...it does include a machine from what I read).

Because it would be a nightmare to support. Apple would have to support every possible configuration of every possible PC model out there. Hardware issues. Driver issues. Peripheral issues. They can do this now because Mac software and Mac hardware are both designed by Apple to go hand in hand. There is no way they could do so otherwise. Microsoft barely manages.

In most cases...not at all. The actual binary code of most applications these days is a fairly small percentage of that 10MB example.

It doesn't exist (publicly) yet. It will likely look like a Mac has always looked (whatever that means)...it won't be ugly I guess.

Well, if you read this guy's thoughts:

Cult of the Mac

it is because the chip that Apple will be using isn't out yet (the Pentium-D). He thinks that this only makes sense in that the content companies are demanding hardware DRM support and currently only recognize Intel's DRM. If we stayed with PowerPC (without content support), we'd be screwed out of copyright material.

I see some logical reasons for Apple's switch to Intel, but the "gut feel" just isn't right. Something seems like a dead-end to me.