Mac is now officially going to be switching to using Intel chips as announced by Steve Jobs at the Apple WWDC Conference today. They dropped PowerPC (IBM) chips.

Any idea of how this will affect audio support on the Mac? Will they be using PC soundcards with updated drivers then?

Also a new version of iTunes was announced, plus Quicktime 7 for Windows should be available for download later today.

Now that is a major change, subject to the intel processor being used (probabbly a standard pentium?) they are little endian compared to big endian of Motoralla / IBM.

Are you planning a Mac (Intel based) version of dBpowerAMP then?

Have to see how this bears out. I bet the mac mini stays with its current processor.

From what I have heard, it will be similar to a Pentium M, but with different instruction sets.

http://www.intel.com/pressroom/archive/rel...0050606corp.htm

The real question for me is, will I ever be able to buy OS X for my existing AMD/Windows PC?
I really don't want to pay the premium for Apple hardware.

my understanding is that they are going to START the transition with the mac mini, then laptops, and finally workstations.

Apple has switched architectures before... about 12 years ago they transitioned from the Motorolla 68000 architecture to the PowerPC line. For a while there they had 68000 emulation, and often 2 sets of binaries with software releases. Messy, but it wasn't awful. This should be similar in complexity to the OS9 -> OSX transition I think.

I'm also 99% sure they will continue to use proprietary hardware, with locked apple-only BIOSes. The fact that this will be an x86 chip is the only real change - don't expect to be able to throw OSX-x86 onto your existing PC or anything.

Nope, x86. The machine they demo was a standard 3.6GHz P4 running OSX.



I don't think this makes much difference at all. They're not switching to windows, so you'll still only be able to use software and drivers written for MacOS. Likewise you'll still only be able to run hardware they sell you. Just now developers will use a different compiler flag, and the 10 people still using (PPC) assembly are out of luck. On the flipside, things like LAME that use x86 assembly will get somewhat faster.

The developer makes the slight changes it takes to allow the program to run on Intel. Steve Jobs claims this does not take much effort. Then XCode will create a universal binary containing binaries for both Intel and PowerPC - supporting old macs as well. Also, an Intel Mac will run PPC binaries through an emulation layer called Rosetta if needed, which has little speed reduction.

The developer for Mathematica for PPC created a working Intel OS X version of the product in 2 hours.

I would seriously consider switching to an apple laptop if it had a fast dual-core intel chip in it. Things I want to see ported to MacOSX now that it is on x86

1. EAC
2. Foobar
3. Monkey's Audio

As you may have read numorous times on the forums. Completely rewriting fb2k may be easier in that case, because fb2k is win32 all over. Do you expect that someone will rewrite fb2k from scratch and then maintain both platforms?

You'll never see that happen. More probable is that someone writes a NEW audioplayer which has some similarity to fb2k.

- Lyx

My two cents worth: if I had a Mac I would be shopping Wintel right now. If the Mac market share dwindles more it is not going to be a very viable platform.

According to steve's speech, Mac marketshare has almost doubled in the prev. 9 months. So, if his numbers are right, macs are booming, not "dwindling".

Apple adheres to the OpenBIOS (IEEE 1275-1994) standard.

I'm a little shocked to find out that Apple is really doing this, after all the rumors.

I'm also disappointed. I realize that if Apple is pretty much unable to get the chips from IBM that they need to bump their laptop line, etc., that this might be their only real option, but from a developers point of view (coming from OS X, not necessarily looking at it from the outside), I think this kind of sucks.

One of the things that has made programming on a Mac so nice for me recently is how elegant the PPC architecture is compared to x86. This isn't really an issue for most desktop apps, but for high performance code, especially where SIMD and assembly is involved, PPC is a whole lot nicer to program for. Altivec has a much nicer instruction set overall, including some things that are simply missing from SSE like the permute unit, not to mention the fact that Altivec has 32 registers vs the 8 registers for SSE. The same goes for general purpose registers. x86-64 seems somewhat underwhelming (especially on the Intel side), and there doesn't seem to be a lot in the future of x86 that will be very groundbreaking (though I suppose they could keep hacking away at it ). In contrast, PPC970 is still pretty new and IBM is still doing really cool stuff with the PPC platform in general as is evidenced by what they're putting into the various next generation video game consoles.

In the end, it probably won't be earth shattering to not have access to all of this in the future, but it definitely makes the Mac platform a little less nice to develop for in the long term. It also makes me a feel a bit like I wasted my time learning hardcore PPC optimization techniques over the last few months Oh well.

I guess the big advantage will be the fact that emulation of windows API's can be done with near native speed now, as opposed to the slow Virtual PC alternative. This should allow access to more of some of the great win32 audio tools in a more practical fashion.

However, the big advantage OS X offers over other operating systems are the extremely polished and very well designed API's that it offers, especially some of the newer ones in Tiger like spotlight, coredata, coreaudio, coreimage, etc., in addition to all of the heritage from NeXT in the form of Foundation/Interface Builder, etc. It's quite true that it should be easier to port programs to Mac OS X now from other OS's on some level (though the biggest advantages will only be seen here for apps containing x86 asm or SIMD), but you still won't get the real OS X experience unless the program is written from scratch to take full advantage of the features the OS offers.

All in all, I'd say that except for some commercial stuff, probably not a whole lot is going to change on the audio front by Apple making this switch.

Agreed on the "not dwindling" part. Besides, almost all the people I know at my university who are doing seriously cool stuff (science or technology wise) are into Mac. Even if for some reason interest were to dwindle, despite signs to the opposite (just look at all the interest from various pc reviewers recently...), there's always going to be a market for some circles.

like this?

That opens the door for Wine type emulation, no requirement for Windows at all. If Apple are smart they will release a version of OSX that can go on any Wintel machine, that is the only way people will try and see, if they can dual boot with what they have. Then they might make inroads.



It is funny but I had the opposite opinion, when recently working extensively with Quicktime it seemed a bit of a mess, perhaps it is just another way of thinking and when it clicks it would be easy.

If Apple want to keep themselves at the computer selling business, they'll never do that. If they did, many people would buy (or just keep) cheap Wintel alternatives instead of buying the more expensive Apple computers.

So, that would have sense only if they converted to a operating systems business and ditched hardware, but I don't see that happening.

It depends on what they want to do. Microsoft didn't become hugely rich and successful from selling PCs, Apple are now in dangerous territory, before hardware was different enough so that they could claim a G4 was better than an Intel xxx and people would believe that and pay a premium for that.

Now they cannot, PCs are a cut throat business and people will be comparing like for like, everything will now be transparent. It wouldn't hurt Apple to have an operating system that people could install onto the 100's of Millions of PCs already in existence and at say $50 - $100 a shot it is money for nothing.

But Apple is not a software company. They have always been mostly a hardware company, and most of the software they make is kind of "basic" software for their hardware platform. Microsoft, on the other side has always been a software company, and sells many other types of software besides operating systems. Also, they became hugely rich because of the huge PC compatible base of computers.

Edit: additions.

No, you're right. Quicktime is a mess. But Quicktime isn't like the other API's that are built from the NeXT foundation to work well with objective-c/cocoa (or the CoreFoundation pure C counterpart). For the most part, it's legacy baggage from pre OS X releases. Some examples of extremely nice OS X derived ones are Foundation, all of the Core* API's, Spotlight, Accelerate (vDSP and all of its relatives), etc.

I would say Quicktime is probably the worst part of OS X, but damn near everything else is very nice. Luckily, Apple is working toward either eliminating Quicktime in the very long run or simply glossing over the gap between it and CoreAudio through the use of higher level audio api's.

Yup..

a) they tried letting mac clones on the market once and lost share rather than growing the market. No reason to think that would change if you can use any generic PC.


Maybe they just want to be a music hardware and software company? Still facing trademark suits from Apple records (maybe McIntosh too?) for moving into music.

They got a good ride from IBM by using baby brothers of the server chips.
If Transmeta (low power x86 compatable processors) hadn't given Intel a big kick in the ass to develop the Pentium M then Jobs wouldn't have such a case of laptop envy.

I really wish Transmeta had been more successful. THeir chips really emulate x86 in firmware and I had the impression that they could have done powerpc as well.

The contents of the Developer Transition Kit shouldn't be taken as an indication of what Apple will be shipping in the end. An Apple kernel developer has even indicated that they are considering EFI. A BIOS strikes me as ancient, inflexible stuff

Looks like project Marklar was real

Speaking as a layman that is not concerned with APIs and Processor optiminzations, if I could run OS X on my current, affordable hardware I would make the switch in a heartbeat.

The only thing that concerns me about the union is the fact that Intel is currently trying to come up with chip based DRM

http://www.macworld.co.uk/news/index.cfm?N...age=1&pagePos=6

That's EFI. It's been around since Itanium was launched, so it's nothing that Intel is trying to come up with. They already did.


Now, for the point about Intel hating Microsoft... I can only see that as insane paranoia. I would understand that they have a certain grudge after MS gave up x86 on the XBox in favour of Power970, but from that grudge attempt to overthrow them? Hohoho.

Also, Apple/MacOS has always belonged in a niche. If Intel really wants to kill Microsoft, funding Linux R&D would probably get them there much faster.




Somebody should tell that guy that he is two months and half late for April first

I think from a business point of view, it is in Intel's best interest to not allow MS to get too powerful or to become to reliant on MS for selling their chips, lest they let MS start dictating processor design to them.

I don't really follow much on the x86 side of things anymore, but as I understand it, Intel pretty much had to implement AMD's x86-64 because of the situation with MS. Maybe this is incorrect, but even if it is, the possibility for something similar still remains.



They already do, or at least they have in the past. I remember a few years ago reading about funding for some distros (at least Redhat) coming from Intel.

This shouldn't come as any real surprise. For as long as I have been looking, there have been references to x86 in bits in pieces in various parts of the Apple Developer documents. In particular, this has been the case for some of the documetation for SIMD related stuff and platform specific optimization. I guess maybe this just wasn't common knowledge.

Oh, I agree. I just don't agree with the wording used in the article. Intel may fear Microsoft's power, but can they really hate the company that helped them become so big?

I can think of several companies Intel hates more than Microsoft.