Apple ditches IBM power PC chip and will use Intel.

[stefankuhn]

whyterabbyt quoth:
and d'you really want to be paying £500 for an OS, Garageband, and next years fashion-parade case?

Yes, I can. (EDIT: Imagine that Mac users would.)

Apple users will and have always been taking the drawback of higher prices for the opportunity to work with an operating system and software they consider to be superiour over anything that is available on x86 right now.

I personally think that this is a very smart move from Apple, and many Mac users will be glad to hear that Apple will finally build Powerbooks and Desktops with 3+ GHz processors. And if you look at what is in store in the future - they'll benefit from intel's hyperthreading technology, as well as multi-core and all the other good stuff.

PowerPC was good when it was young, but the technology is now at a point where it can no longer provide what is demanded in future applications. I was absolutely delighted when I heard Steve Jobs keynotes. And I'll be affected by this transition, much like any other developer that wants to / has to develop cross platform.

In the long run we will no longer worry about processor specifics like little/big endian, vector engines, and stuff. So it's good news not only for users, but also for us devs.

Plus, the tools for doing the shift are said to be very easy to use. I very much hope so, but I have no reason to not believe it either.

Cheers,
Stefan

[Borogove]

1) Altivec optimized stuff will a lot harder to port. Which is most audio stuff. My guess is emulation of altivec instructions won't be fast enough.

It should be straightforward but tedious to port Altivec to MMX/SSE/SSE2. However, due to Altivec having more & wider vector registers, the ports will be inherently slower, spilling to memory more often. Emulation of Altivec may be slower than emulating a straight C version, even.

3) Porting VST plug-ins that use VST GUI will be close to trivial. This means that possibly devs like ConcreteFX, RGC, and Big Tick might port their plug-ins.

Unless a substantial amount of the plugin was in assembly to begin with, Apple using x86 won't make a substantial difference here. Personally, the development environment differences are a bigger hurdle preventing me from building Mac versions of my plugins than the CPU architecture.

4) Apple will probably use a custom motherboard with their own chipset as opposed to a standard chipset. (Schiller) stated that they will actively prevent the Mac OS from running on non Apple PCs.

If nothing else, we can assume that all the hardware device I/O interfaces will be different from their PC counterparts. Consider, just to pick one thing, the I/O map of the USB controller. In the PC world, you have two dozen Taiwanese manufacturers who all want their stuff to interoperate, and all are happy to have Linux work on their stuff as long as they don't have to do any work to support it themselves. It's easy to find out where the I/O ports are and how they're supposed to be used. On the Mac side, you would have to reverse-engineer the OSX-on-Mac device drivers from binary, then write all-new drivers for OSX-on-Standard-PC. It would be like turning back the clock to the situation on Linux in the early 90s, only with an active opponent instead of mere indifference.

[shamann]

I cant see the Mac audience being amazingly happy if they find that they're getting sold a box which ties them to DRM...

But the iTunes audience isn't the same as the Mac audience. Think about a MacMini in place of your VCR/DVD player, attached to your television as a nothing but a movie storage device/player, with some kind of powerful airport attachment to do the real work. The DRM is more important there, because the movie biz folk won't facilitate retail film downloads without it.

[stefankuhn]

I cant see the Mac audience being amazingly happy if they find that they're getting sold a box which ties them to DRM...

Crap.

DRM is already an integral part of ITunes. Why should any mac user bother?
Mac users buy music at the ITunes music shop and listen to it on their box or IPod. And they'll only do this quicker on an intel based machine.

If you meant to mention the TCPA / palladium chips that are on most modern mainboards, noone said that Apple is going to use them. However, I don't see any reason why they shouldn't.

Cheers,
Stefan

[slasi]

slasi quoth
4) Apple will probably use a custom motherboard with their own chipset as opposed to a standard chipset. This will make it harder for people to hack the Mac OS to run on a generic Intel PC. The open source Darwin kernel might make it a bit easier for hackers.

I really doubt it, to be honest. I expect to see them using standard Intel chipsets, even if the mobo's are their own. I dont think Apple have got the expertise that's needed.

Uhm.. Apple have been developing their own motherboard chipsets now for several decades, so they definitley have the expertise. Although I do agree that it is probaly just easier/more efficient for them to use Intel's chipsets. So most likely their motherboard will be custom with some unique features.

6) In the long run, update hassles and plummeting resale values aside, this will be good for most Mac users. There will no real difference in the way their
Macs behave, other than being more current with processor speed upgrades. Not sure if Apple will pass along any of the cost savings (it will probably have to, to some extent).

This is contradicted by your point (4). If its custom motherboards and chipsets, yes there will be a 'difference in the way their Macs behave'. And keeping up development of your own chipsets for Intel processors is going to be an expensive business. That would rule out those cost savings. Plus if the make their own chipsets and they arent up to the standards of Intel, VIA or NVidia, the move to x86 is going to make their kit look foolishly dated. Thats already a criticism of the current Mac's, and it would be a real blow if Apple couldnt get out those little 'technology leads' that they've managed in the past (first with mainstream USB, Firewire, DVD et.c.)

I meant no different in the way the Mac behaves with regard to their current PowerPC Macs, not PCs.

[Borogove]

slasi quoth
4) Apple will probably use a custom motherboard with their own chipset as opposed to a standard chipset. This will make it harder for people to hack the Mac OS to run on a generic Intel PC. The open source Darwin kernel might make it a bit easier for hackers.

I really doubt it, to be honest. I expect to see them using standard Intel chipsets, even if the mobo's are their own. I dont think Apple have got the expertise that's needed.

I disagree. Apple has been building highly advanced motherboards on their own for quite some time, and a non-standard chipset is their best weapon in the fight to keep OSX running only on Mac hardware. Okay, maybe their second best weapon after their legal department.

[Jeremy_NSL]

This whole thing is sort of humourous, considering Apple's marketing stand on the G5. "Supercomputer in a box!" "Destroys any Intel processor!" followed by graph after misleading graph.

Steve Jobs is now like, "We need a faster processor". No shit Steve! Everyone who was able to look past the marketing has known that for 10 years - since the beginning of the PowerPC transition!

But of course its about more than just cpu speed. I agree with the Wired reporter that the main reason for the switch is DRM for audio/video downloads. And of course it can't just be cpu speed, because Apple has always been slower and more expensive and Jobs never cared before.

[slasi]

3) Porting VST plug-ins that use VST GUI will be close to trivial. This means that possibly devs like ConcreteFX, RGC, and Big Tick might port their plug-ins.

Unless a substantial amount of the plugin was in assembly to begin with, Apple using x86 won't make a substantial difference here. Personally, the development environment differences are a bigger hurdle preventing me from building Mac versions of my plugins than the CPU architecture.
indifference.

I'm basing this on the assumption that not having to rewrite the assembly code will be significantly less work than having to rewrite for a new architecture. I hadn't considered that the difference in development environments and assembler mnemonics in particular might still be a significant barrier.

[whyterabbyt]

stefankuhn quoth

Yes, I can. (EDIT: Imagine that Mac users would.)

Apple users will and have always been taking the drawback of higher prices for the opportunity to work with an operating system and software they consider to be superiour over anything that is available on x86 right now.

Hmmmm. I certainly see a lot of 'legacy' Mac places who have stayed primarily because of specific applications. I think Apple recognise, hence things like FCP and Logic; they want to ensure that application-based audience. As soon as it turns out that these apps 'will run on PCs as well' (whether they actually do or not it'll seem the same to less technical folk) then I think people will wonder why Apple are hamstringing them for £500 extra a machine.

My department is primarily Mac based, and I run the only group of Wintel machines we have. But the folk who actually have to pay out the money are getting tired of forking out twice as much per seat only to see it outgunned by the commodity PC's I go for. Lots of people around me are starting to say 'umm, but why are we paying twice as much[/i] for every machine. Its not the actual operating system they want to use, its an 'industry standard' video editor, or sound editor, or animation package.

[whyterabbyt]

slasi quoth

Uhm.. Apple have been developing their own motherboard chipsets now for several decades, so they definitley have the expertise.

They have what you might call 'generic' expertise , but not necessarily specific to the Intel processors. And I consider it highly doubtful that they can compete head-on with Via, Intel and NVidia from out of the blue. There's a lot of money and expertise in those chipset companies, and Apple's stuff has lagged behind the cutting edge for some time now.

Although I do agree that it is probaly just easier/more efficient for them to use Intel's chipsets. So most likely their motherboard will be custom with some unique features.

That seems far more likely to me.

I meant no different in the way the Mac behaves with regard to their current PowerPC Macs, not PCs.

My apologies.

[whyterabbyt]

Borogove quoth

I disagree. Apple has been building highly advanced motherboards on their own for quite some time

And are now totally outstripped on that front by the PC manufacturers.

and a non-standard chipset is their best weapon in the fight to keep OSX running only on Mac hardware. Okay, maybe their second best weapon after their legal department.

If they do it, that'll be why. Not because they can make better chipsets. If they could, they would be doing by now. But their development cycle is too slow, and its hitting the way it did the big Unix Workstation vendors like SGI.

[snooky]

amd has the only sensible chips on the market..the rest are just toys; toys for kids with SBlives and FlStUdiO!1!!1

[Charlie]

I run a Design Studio where we use Macs and PCs. We are client led and have to use the software the clients use. Having worked for the past ten years on both systems I would like to agree with whyterabbyt that it is the application software and file formats that lead the market. The transition to OSX was expensive and I am not keen to spend the money again, then again I have no choice. The biggest winner here is Intel.

I would rather buy new PCs and cross-grade software than buy new macs and upgrade software. Whatever happens it will be interesting to see how Apple make it work and keep their users on board.

(The irony of working with both Apple and MS systems is that you realise that they are both essentially the same, much like Logic and Cubase are the same. I'm not loyal to either.)

[stefankuhn]

As soon as it turns out that these apps 'will run on PCs as well' (whether they actually do or not it'll seem the same to less technical folk) then I think people will wonder why Apple are hamstringing them for £500 extra a machine.

Wait a second. I think you are making wrong assumptions here. PowerPC != Mac Computer and Pentium != PC.

A mac (or windows pc) does not define itself by the processor used. Windows users are mostly more willing to accept that windows is not equal pentium since you see windows also on ARM and Dec Alpha processors. For Mac, PowerPC may have become a synonym for the operating system ever since the transition from 68K to PPC.

I'm sure you're well aware of all this, but don't expect Apple to allow dual-boot systems with windows on any Intel hardware not manufactured by Apple. MacOS X will never be available for DELL or self-built PCs.

You need to look at it from this angle: It's just the same Mac, it will look the same and it will run the same software and behave the same. The fact whether it runs on Intel or PPC processor will never be of notice for the user. And never be of interest.

At least that is the idea.

Cheers,
Stefan

[stefankuhn]

I run a Design Studio where we use Macs and PCs. We are client led and have to use the software the clients use. Having worked for the past ten years on both systems I would like to agree with whyterabbyt that it is the application software and file formats that lead the market. The transition to OSX was expensive and I am not keen to spend the money again, then again I have no choice. The biggest winner here is Intel.

I would rather buy new PCs and cross-grade software than buy new macs and upgrade software. Whatever happens it will be interesting to see how Apple make it work and keep their users on board.

You should see it as an option, not an either-or-case. PPC and Intel machines will co-exist at least for the next years. If your office doesn't need new, faster computers, you stick with your PPCs and still can benefit from all new OS and Application releases, because Applications will be built for both architectures.

There is no obligation to update, it's a lot easier than the shift from OS 9 to OS X, where OS 9 was finally discontinued.

Best,
Stefan