Does it matter? Is it a big deal? Does it mean that Apple will be able to compete Apples to Apples with their competitors now? Will their be lawsuits as 13 year olds, hack homemade boxes to run bootlegged copies of OSX intel? Any predictions?
Maybe it’ll finally bring their hardware prices down to within a reasonable (IMHO) range. Then again, when you buy a Mac these days, you’re still mostly paying for the branding and packaging, anyway.
If it turns out OSX’ll run on the A64/Opteron, even better. Actually, that’s where I’d prefer they concentrate their development efforts, along with the Pentium M, instead of the P4. That’s where the price/performance sweet spot is right now.
It’d be even better if they could beat out MS with a 64-bit x86 commercial consumer OS.
Microsoft already has a 64-bit version of XP running on Opteron.
Yeah, I know it’s running. I meant a consumer-oriented boxed release, eg, XP-Home-64.
Something I wrote elsewhere on da intarnet:
…The decision was apparently made because IBM/Freescale hasn’t been able to deliver the PowerPC performance improvements they had promised. Unofficial scuttlebutt is that IBM has been deliberately dragging its heels on delivering more and better PowerPC processors to Apple, since Apple’s XServe server line competes directly with IBM’s high-end businesses. In any event, Apple (read: Steve Jobs) has apparently had enough; the Intel alternative has always been kept as an emergency alternative, and now Jobs is invoking it.
Frankly, I’m slightly surprised at how much a-twittering the geek community has generated over this. From a business and technological point-of-view, it makes perfect sense. The biggest reasons (in the past) for preferring PowerPC over Intel were performance and heat – both of which are moot, especially if Intel does deliver the low-power consumption CPUs they’ve promised. Given Steve Jobs’ well-reknown streak of perfectionism, there’s no reason to believe an Intel-based Mac will be any hotter or noisier than what Apple is already selling.
And from the end-user’s point of view, this is no big deal. The biggest reasons for preferring Apple computers over Windows-based PCs have been superior hardware/software integration and ease of use, and this hasn’t changed, either. The use of “Rosetta” and “Universal binaries” are major benefits here, since that means an Intel-based Mac will be able to use most of the existing PowerPC software that’s already out there. Apple’s switch from 680x0 processors to PowerPC was a more complex transition than this (raise your hand if you remember “fat binaries”).
Now, granted, the devil is always in the details, and the transition from PowerPC to Intel won’t be perfectly effortless (why else is Jobs targeting a switchover by 2007?). But that’s why the news was announced at a developer’s conference; from the perspective of the non-geek users out there, this whole thing is much ado about nothing.
Someone at Intel is getting a big fat bonus cheque for this marketing coup, but otherwise I agree with rjung. Macs make up about 3-5% of the market for PCs, so this isn’t gong to be a huge windfall for INTC, and the end user will hardly notice a difference. But it makes good news!
Good post, Rjung. Very well thought out. Thanks.
What about the rumors that the reason Apple is moving to Intel is because by doing so they will be able to integrate DRM into the hardware?
The -D series of Intel Pentiums does not have the DRM it was rumored to have.
I had a link in the other thread.
Second: This means WINE will, or should, run with a recompile, on all Macs. This means most video games will be Mac-able. This is a huge win for the Unix side.
Looks like you have a new sig, Rjung. You’d have to leave out a link, of course.
I don’t dare look at what the Mac true believers have to say.
You mean rjung isn’t one?
No matter how annoying he gets, he NEVER gets THAT annoying!
I was at one board where they were discussing how quickly Intel could get PPCs, and I assume the rest of the chipset, into production. I didn’t post a correction, that Jobs was talking about using Pentia, because when you are that far gone in Mac addiction the reality you have constructed can be rather fragile. But since I signed up there because I’m thinking of building an Apple I repro with my daughter I’m nobody to throw stones.
Here’s a typical thread in re the reaction of the Mac community. Surprisingly good signal-to-noise ratio.
I think Apple’s stock might take a (hopefully temporary) tumble, since perhaps many folks looking at getting a new G5 might hold off until this new hardware is available, or at least until the dust clears. But I think this will ultimately work out well. Think I might buy me some shares of APPL soon . . .
<Raises hand> (I also recall building fat PPC Mac binaries using an IBM minicomputer; it took 9 hours to build the Mac app I worked on with full optimization.)
I don’t see that this transition as fundamentally different than the switch from 68k to PowerPC. Both featured a switch in processors, an emulator for old code, and both will have fat binaries (except Apple now calls them “Universal Binaries”) where old and new code will be compiled into one app.
This creates a somewhat simple situation for end-users: old PPC apps will still run, just not optimally. New Fat or Intel-only versions of apps will slowly be brought forth as developers finish them
For developers, though, it’s a huge pain in the ass: you have to build the app twice and test it twice. Productivity drops by half for the testers. Twice as many test machines to test on. etc. etc. This is exactly what happened during the PPC migration. It usually worked OK; sometimes it didn’t as you found bugs in the emulator or differences between compilers or what have you.
It seems like this will be a huge problem for Apple’s machine sales for the next 6-12 months: who the heck is going to buy a G4 or G5 Mac, knowing it will be obsolete very soon?
Very little time to reply, I’m afraid…
I’m quite surprised at the lack of mass suicides in the Mac community. It’s really too soon to say whether this will be a good or a bad move, and we’ve simply no idea how well future offerings from FreeScale and IBM might have stacked up against Intel’s purported pipeline. We know virtually nothing about how an Intel-based Mac will designed, how commoditized it will be, whether AMD processors will ever work in Macs (I sure hope they can, but it wouldn’t surprise me if some of Intel’s new license technology precludes the MacOS from running on anything but an Intel chip), etc.
What’s odd to me is that, when one really gets down to the real numbers, cost isn’t much of an advantage, nor is power consumption on the PPC end really so bad. The 970FX actually compares pretty well with the Pentium M, given IBM’s PowerTune technology, and FreeScale does have some pretty stellar dual-core PPC processors ready to ship late this year or early next, if they stick to schedule (a big, big if, I’ll grant). IBM was selling Apple chips practically at cost. Can Intel really do better? What is this fantastic pipline Intel is supposed to have? Does it really compare that much more favorably to what IBM can offer? IBM will, after all, be supplying what are essentially triple-core G5’s to Microsoft by Christmas running at 3.2GHz, and given the size of the box, it can’t be all that wattage-hungery or the case would melt. The CELL processor over on Sony’s side looks incredible. The three major game consolers picked IBM for a good reason: Better price-performance and a more advanced architecture than anything on the x86 side.
And Apple has been touting the superiority of PowerPC for years. The 970 did hit a wall at 90nm, but so did everyone else’s processors. What gives?
I’m guessing IBM could have delivered more for Apple if they wanted, and probably put the 970 and its variants on something of a back-burner because they can make a hell of a lot more money off of Microsoft and Sony with their respective consoles. Apple was probably more of a nuisance than anything, in this market environment; and it’s no secret Apple only makes up about 1% of IBM’s sales.
They won’t make a bigger difference to Intel’s bottom line, either. Neither chipmaker needs Apple. Apple desperately needs processors that can compete. For reasons I can’t quite fathom, Intel seems intent on delivering that enthusiastically to Apple. Why Apple should matter at all to Intel, I just can’t understand, unless they feel somehow threatened by Microsoft’s abandonment of Intel processors for the Xbox360. It’s really a fascinating market that’s developing, with some very strange bedfellows.
Now, at the very least, the “MHz myth” is history. For the average user, the chip will make absolutely zero difference, buried as it will be in whatever snazzy box Apple ships. I’m a little bummed, as I would have liked to see what the PPC platform could have delivered in the future; but not that much.
I’m guessing there’s a whole lot of politics, seething rage, and heaping doses of misinformation behind today’s keynote, but that’s always been the backdrop of any major Apple transition, so nothing really has changed. My guess is some people at IBM are happy to be rid of a customer like Jobs. And clearly, Jobs is plenty happy to be rid of the folks who made him feel he could promise a 3.0GHz G5s a year after his 2003 keynote, only to have 2.7GHz G5s today. It wasn’t that long ago that the G5 was Apple’s savior. Not that long ago at all. The fact OSX on Intel has been in development for 5 years makes me wonder how realistic the hype back then really was. Figure Microsoft has got Longhorn for PPC builds lurking in Redmond somewhere? I’m guessing absolutely not. I guess you can’t believe what any of these folks say; so all the more reason to sit tight and see what the future brings. I withold all judgement at this point on whether it’s a good move or not. There are simply too many good arguments either way, that will only be settled when we see some actual products.
I’ve seen responses all over the board.
Frankly, I think the folks who are saying they’ll drop the Mac after this are just a bunch of wanna-be drama queens. What, like they’re going to go to Linux or something?
I think it depends on what you’re writing. If your app deals primarily with high-level stuff, you probably won’t need to do anything more than to check an extra processor. If you’re doing low-level stuff, however, you have to add additional checks for architecture and networking differences (endian being the most obvious example).
I agree that Apple’s hardware sales will take a hit, but I think they have enough cash in the bank to weather it out.
That said, there’s no reason to believe Apple will stop supporting PowerPC-based Macs by the time they’re fully switched to MacIntel (Jobs’ term) gear. There will still be a sizable base of existing PowerPC users out there to sell to, and Apple would be foolish to pass up the chance to sell them new versions of MacOS X/iLife/iWork/etc.
My WAG is that if Apple transitions completely to Intel by 2007, expect them to drop PowerPC support by 2009-2010 at the earliest. So the computer you buy today is good for at least 4 or 5 years, which isn’t too shabby. Then again, mine is already five years and going, so…
As you noted, with Microsoft and Sony and Nintendo going towards PowerPC processors for their next-gen game consoles, IBM has no incentive to cater towards Apple any further. Corroborating this is a report in the New York Times that says IBM was barely breaking even in supplying PowerPC CPUs to Apple before; when Apple went into their latest negotiations and wanted a price cut, IBM said no.
IBM will not shed any tears over losing Apple’s business, that’s for sure.
One technical advantage of this switch is that it probably means Apple will have an easier time getting cutting-edge graphics for their MacIntels, since the endian difference is IIRC the biggest obstacle currently preventing from having their choice of the newest graphics cards.
I suspect IBM made promises to Apple, Jobs used those promises to make his public announcements, and IBM dropped the ball, leaving Jobs with egg on his face.
Frankly, I’m more surprised that Jobs publically admitted to his 3.0GHz gaffe today than anything else… :eek:
http://www.theinquirer.net/?article=23708 is the DRM article I meant.