“Did Apple make a mistake by switching to Intel? We may never know, but Apple had more options than has been previously reported,” Ashlee Vance exclusively reports for The Register. “A chip start-up that created a high performance, low power processor compatible with existing Mac software had been working closely with the computer company for many months.”
“Apple was looking for a new chip supplier largely because it was struggling to find a decent part for its key laptop line. IBM could not deliver the right performance per watt characteristics needed for slim, powerful kit and was struggling to produce chips as efficiently as Apple would like,” Vance reports. “PA Semi – a maker of low-power Power processors – formed a tight relationship with Apple – one meant to result in it delivering chips for Apple’s notebook line and possibly desktops. The two companies shared software engineering work, trying to see how Apple’s applications could be ported onto PA Semi’s silicon. When word leaked out that Apple had signed on with Intel, it shocked the PA Semi staff, according to multiple sources. ‘PA Semi was counting on that deal,’ said one source. ‘They had lots of guys walking around in a daze when Apple went to Intel. They had no idea that would actually happen.'”
Full article here.
[Thanks to MacDailyNews Reader “Chas” for the heads up.]
MacDailyNews Take: Apple made the right choice.
Advertisements:
• Introducing the super-fast, blogging, podcasting, do-everything-out-of-the-box MacBook. Starting at just $1099
• Get the new iMac with Intel Core Duo for as low as $31 A MONTH with Free shipping!
• Get the MacBook Pro with Intel Core Duo for as low as $47 A MONTH with Free Shipping!
• Apple’s new Mac mini. Intel Core, up to 4 times faster. Starting at just $599. Free shipping.
• iPod. 15,000 songs. 25,000 photos. 150 hours of video. The new iPod. 30GB and 60GB models start at just $299. Free shipping.
• Connect iPod to your television set with the iPod AV Cable. Just $19.
• iPod Radio Remote. Listen to FM radio on your iPod and control everything with a convenient wired remote. Just $49.
Related articles:
Apple CEO Steve Jobs’ ultimate goal: ‘to take back the computer business from Microsoft’ – June 16, 2005
The Mac platform does NOT need to become like WIndows with MILLIONS of useless apps that people think is such a big advantage – it isn’t an advantage if they’re all mediacre rubbish apps.
I’ll have 50 EXCELLENT OSX apps anyday compared to 5 million crappy Windows ones anyday thank you.
The only person/team able to judge what was best for Apple was Jobs/Apple management. So let them get on with it and show what they can do without tearing at them with ridiculous speculation..
..and so far their efforts on Intel ain’t half bad. My MBP is a blinder.
But no doubt I’ll be pissed off by the end of the year when there are newer faster, cooler ones out..
Apple went intel to become part of the windows world. they must run windows in order to be accepted as part of the windows world.
in 5 years most software will be windows only; the only company making apple software will be apple.
UB gives Apple complete control over their chip selection. IF something better than Intel comes along, I’m sure Apple will consider going that route.
Re: “in 5 years most software will be windows only;”
It already is. So. As long as Apple is still motivated to create OSX, iLife, etc. It will survive.
Lots of people, including MDN, seem to have the opinion that Apple made the right choice. They may have, but do people really know enough about this PA Semi company? I certainly don’t. Hard to have an informed opinion without all the facts here. (Of course, I haven’t read the whole article though, Maybe that would have helped?)
Whine whine whine whine whine whine whine. PPC has been replaced by Intel, oh woe is us. All this whinging and kvetching seems so familiar, somehow.
Oh, right! I remember now! This is exactly the same thing we went through when the Mac switched from OS9 to OSX! Remember all the wailing, mournful “The Mac is dead” posts?
Thankfully, we got over that transition and accepted it. The same thing will happen with Intel.
(Note: I guess not everybody has accepted the OSX transition. The web site thalo.net is home to a group of malcontents who have been bitching about OSX since 2001. They began on the MacFixIt forums, then set up their own page when MFI closed their board. The site is a fascinating case study in losers unable to get on with their lives. Five years, four major OS revisions, and these pathetic whiners can’t let it go.)
The only thing special about an (intel) apple computer is that its easier to get it to run osx. Otherwise its the same thing the other guys are shilling. Originally I thought maybe apple would have better build quality but 5 revisions to the mac book pro killed that idea. Oh but wait…theres a camera.
I am going to hope this article is bs, and that apple really never has a choice anyway.
you asstards. you all cursed intel chips when apple used powerpc chips.
now you love intel chips because apple switched to them.
what a bunch of hypocrites.
i’d still rather have a dual AMD opteron apple system…
MDN doesn’t have enough technical expertise to make such a claim. Leave that for the engineers, not the “journalists”,
As for PA Semi being shocked? As hard as it is, that’s the risk you take as a start-up. Happens all the time. Big company gets small company to devote their time, energy, and equity to a product devoted entirely to the big company. Big company has multiple options. Big company makes decision – and not for the small guy. Big company doesn’t really care about the small guy.
That’s life in the start-up world. Don’t work for one if you can’t understand that risk.
I love how the crusty old Mac users are still unhappy about the switch to Intel. With Intel we’re now getting:
– Incredible performance gains for laptop and consumer Mac users
– Nice performance gains for G5 workstation users when Core 2 Duo comes out in a couple of months
– The ability to run Windows at near-native speed via Parallels
– The ability to boot into Windows to play games and whatever else made potential Mac users hesitate
– Performance and price parity with the x86 world
It’s painfully obvious that the Intel switch was the way to go. I had given up on Apple and sold my PowerBook since it was so pathetically slow compared to Windows laptops, but I’m happy to be using a Mac again with my MacBook Pro.
The 64bit, 4MB L2 cache, 128bit SSE4 vector processing unit equipped Core 2 Duo is really going to put all of the skeptical people at ease. Just 2 or 3 more months until it replaces the G5 workstations, and another month or two after that until it gives MacBook Pros a big fat upgrade.
Apple had a hard enough time getting speed and efficiency improvements from big chipmakers Motorola and IBM, so now it was supposed to rely upon design specs, a song and a prayer from PA Semi?
Please.
Apple needed faster, lower-power laptop chips. Don’t forget, the iMac and Mac Mini basically use laptop chips, so most of Apple’s offerrings need chips that don’t require a separate A/C unit to cool. That leaves out anything IBM was going to produce (cell, G5, etc.), FreeScale (G4 incremental improvements, not quick enough to keep up with Intel and AMD), and PA Semi (who wasn’t going to have any sizable volume until 2007).
Yes, 2007 was the target date for volume production from PA Semi. That’s still at least 6 months away, and that’s assuming the target date was met (and we all know about target dates with the PowerPC architecture).
Apple didn’t just make the right choice – Apple had no “choice”, because there was no real alternative.
In my heart I would hope that this – finally – lays to rest the BS that Apple ‘had no choice’ … but I’ve been reading on this site for far too long. That’ll never happen. The mental retreads are already poo-pooing this verification; ‘Who’s PA Semi?’ ‘How do we know these guys were the real deal?’ blah blah blah.
From the article: “PA Semi secured a large amount of venture funding due in part to the stellar technical reputation of its staff. Former Digital alumni include VP of architecture Peter Bannon … Leo Joseph, the COO and Jim Keller, the VP of engineering. Several of these engineers did much of the key work behind DEC’s Alpha chip, which for much of the 1990s was consistently the fastest microprocessor on the market…”
Phhfft! What a bunch of amatuers! Right guys?
“PA Semi’s first processor – the PA6T-1682M – is due to sample in the third quarter of 2006 as a 2GHz, dual-core product with two DDR2 memory controllers, 2MB of L2 cache, and support for eight PCI Express. The product will ship in volume next year and be followed by single-core and quad-core chips. It also supports the Altivec floating point instruction set that currently provides a massive speedup for multimedia and scientific Mac software. At 2GHz, the chip consumes just 7 watts of power according to PA. Intel’s Core Duo consumes between 21 and 25 watts.”
Lessee; Apple could have had, by this September, a laptop chip that uses less than HALF the power of Intel’s vaunted Core. In the meantime they could’ve been using the dual core G4 from Freescale to tie them over, which ALSO uses less power than Core. And with IBM waiting in the wings, eagerly trying to get Apple to evaluate and advise on it’s needs for either the Hollywood CPU (Nintendo’s Wii) for a home multimedia device, and/or the upcoming Power6 (even better) for the desktop/server space, how can any one HONESTLY say that Apple wasn’t absolutely in the CPU catbird’s seat BEFORE switching to Intel?
God god – for the first time in the company’s history Apple had its pick of a wide variety of excellent CPUs, from 3 different vendors, one of which was clearly willing to do Jobs’ bidding for the forseeable future just in return for getting it’s foot in the door!
How many different ways can you spell “idiot”? Well, I know one way that should make it into the dictionary ASAP – we should spell it J-O-B-S. His last minute blow-up has left us with melting MacBooks, and wholly beholden to a CPU vendor that is clearly on the downslide vs it’s only competitor (AMD).
Don’t believe it? Just ask DELL.
Tell me, Odyssey, since it’s clear that you didn’t bother to read any of the other comments in this thread, many of which rebut your points, why should I waste my time reading your long-winded tripe?
umm … because I’m cute?
Oh, and just for the record – disagreeing and rebutting (even if unsuccessful) are two different things.
The Mac platform does NOT need to become like WIndows with MILLIONS of useless apps that people think is such a big advantage – it isn’t an advantage if they’re all mediacre rubbish apps.
I’ll have 50 EXCELLENT OSX apps anyday compared to 5 million crappy Windows ones anyday thank you.
Problem is, it’s usually just ONE of those 50 million that you happen to need, and it happens to not run on a Mac.
People think that the Mac’s 50 great apps are better than the kajillion a PC has… but there are a LOT of specialty apps that have been written for the PC, especially in the business world… which is why Macs will NEVER make inroads into the corporate sphere.
Odyssey67:
Even if Apple could have shipped a PowerBook using the “PA6T-1682M … a 2GHz, dual-core product with two DDR2 memory controllers, 2MB of L2 cache, and support for eight PCI Express…”, it wouldn’t have shipped until 2007.
The current road map gets us a 2.53Ghz Core 2 Duo MacBook Pro by then, with 4MB of L2 cache and 64bit support. It will utterly kick the crap out of the CPU you are describing.
Not to mention the 1 year+ of Mac users enjoying the Core Duo in the mean time, the native x86 support for booting into Windows or running it via Parallels, the lower prices…
What is it that is really bothering you?
Nick says: “Even if Apple could have shipped a PowerBook using the “PA6T-1682M …” it wouldn’t have shipped until 2007.”
As I said above, Freescale’s had a dual core G4 available for about a year and could have plugged that hole. Like many people, you’re assuming, just b/c Apple didn’t use Freescale’s better G4 CPUs, that they weren’t available. Or that they weren’t good performers. Apple didn’t use them b/c they wanted to put their upcoming Intel transition in the best light possible. In fact, most of the CPU only benchmarks I’ve seen, that tested a single core Core CPU and the old 7xxx G4 PBs, showed the latter losing (of course), but not by much. The new e600 core G4s (single or dual), at the same clockspeeds as the Yonah chips, and with all the other system advantages of PCIe and Serial ATA, would almost certainly have matched what we see with the MacBook bros now (and used less power, btw):
http://www.freescale.com/webapp/sps/site/overview.jsp?nodeId=0162468rH3bTdG7249
When the 64bit PA6T was ready, the Freescale parts could have been moved down to iBook/MacMini territiory. No muss, no fuss, no software transition or price increases required.
“The current road map gets us a 2.53Ghz Core 2 Duo MacBook Pro by then, with 4MB of L2 cache and 64bit support. It will utterly kick the crap out of the CPU you are describing.”
That’s not a realistic conclusion. First, the PASemi part, like all CPU designs, has it’s introductory speed rating, and then a certain amount of overhead built in, to allow for product upgrades throughout the life of the design. Second, your faith in Intel roadmaps is stronger than most. The current one has morphed about three times since Apple came aboard. Plus, Intel has a track record of delayed rollouts, so MAYBE we’ll see 2.53Ghz in early 07 – we’ll have to wait and see. And last, since the guys designing the PASemi CPU are the same responsible for the legendary Alpha chip, I’ll put my money on them in any ‘crap kicking’ contest with the corporate sloth Intel has become.
“Not to mention the 1 year+ of Mac users enjoying the Core Duo in the mean time, the native x86 support for booting into Windows or running it via Parallels, the lower prices…”
I’m not “enjoying” a new Mac because I use Photoshop. The 1Ghz iBook runs it faster than any Macintel can right now. So I have to wait a year regardless. Booting into Windows? Parallels? I’ll pass on the insecurity myself, thanks, though I’m sure some geeks will love it. A safer alternative would be in using a Hypervisor kernal, which some think Leopard will have. However, that’s actually done more easily with a PPC/RISC CPU:
http://macdailynews.com/index.php/weblog/comments/9603/
(click MDN’s Wikipedia entry for the RISC scoop)
“What is it that is really bothering you?”
I’d think that would be obvious, based on what I’ve already written: The amount of time and money that’s needlessly been wasted – by Apple as well as their software developers – on this transition. I guess you can include us buyers too, now that every Macintel is about $100 more expensive than its predecessor.
The more one looks objectively at what’s unfolded, and at the circumstances Apple was in before they undertook the switch (such as PA Semi bending over backwards for Jobs, to no avail), the more one has to accept that this ‘trip’ was not necessary. UNLESS you think DRM’d/TPM’d computers, locking down all things ‘multimedia’ that you own (if the concept of ‘ownership’ even survives such technologies), are worth it. Personally, I don’t.
And yet, even though selling video was what drove the Macintel decision, Apple’s finding it hard to devote the time and resources towards making what everyone wants from them – a robust, elegant, and dedicated multimedia hardware offering. Intel apparently can’t produce enough chips fast enough (deja vu), and Apple’s too busy messing with the bugs in their new Macintel products to focus on it. Meanwhile a 40% sales increase in their computer market (2004-05) has all but stalled according to recent reports.
I hope I’m being clear now.
” width=”19″ height=”19″ alt=”cool smile” style=”border:0;” />
Odyssey67:
“I’m not “enjoying” a new Mac because I use Photoshop. “
Ahhhh, ok, I see what you’re really upset about now. The blame for that problem, as you probably already know (you seem well-informed), is that Adobe was still using Metroworks CodeWarrior for Their Mac software. Everyone else who was using XCode had a really easy time transitioning to a universal binary, but Adobe pretty much screwed their customer base (people who paid good money for their Adobe software like you) by failing to move to XCode in a reasonable time frame.
I think that this will also be a “good thing” in the long run, since once everything is in universal binary, it will also be in a pure Cocoa application state. We’ll finally be rid of all of this Carbon crap.
“most of the CPU only benchmarks I’ve seen, that tested a single core Core CPU and the old 7xxx G4 PBs, showed the latter losing (of course), but not by much”
In Cinebench, doing a pure floating-point CPU scene render, my old 1.5Ghz PowerBook scored a 135, and my MacBook Pro 1.83Ghz scores a 264 when I disable my second core via the system pref option that gets installed along with XCode. In other words, I see 96% higher performance using a 22% faster CPU.
The Core Duo generally crushes the G4. The only time it doesn’t is when you’ve got an application which is assembly-optimized to make good use of Velocity Engine – but once Core 2 Duo ships with SSE4 we’ll have a decent Velocity Engine substitute.
Nick:
I’m actually a fan of the whole concept of UB, and think the freedom it gives Apple AND software developers is great – a true innovation. And no, I don’t give Adobe any slack for not going with it sooner – no matter their reasoning (huge software base) it was long overdue for them to begin the move. However, I would caution that you not read too much into that one small aspect of my argument. I only brought up Photoshop to demonstrate that a large portion of Apple’s market can’t truely “enjoy” much about the Switch.
As for the Cinibench scores, I’m not disputing them – I just think they aren’t too bad for a 2 generation old version of a CPU with 1/4 the cache. My main point was that the G4s Apple used in the last PBs, from which these scores come, were not the best available – even in single core form. The fact is, 2x larger caches, 4x faster FSB speeds – basically everything that makes Yonah so good – are all realities on the e600 core. I’ll say it again; I don’t know whether the e600 G4 would be equal to Yonah in all respects, or any respect – we’ll apparently never know the answer to that, b/c Apple declined to use the chips when they could have. What I do know is that they would have been a substantial boost when compared to the old G4 they were using (and all the Macolytes would have been singing it’s praises as a result), AND that it would have been more than enough to ‘hold the fort’ until the 64bit PASemi part came on line.
Once that fact is conceded, and objectively it has to be, then the whole ‘x86 is Good’ argument swings towards how the world can now run Windows on an Apple-branded machine. Yet, all this expectation about how great running Windows stuff (apps or OS) on a Mac is, ignores a] the marginal utility it actually provides most people, as well as b] the kernal-level technology that’s existed for a while (Hypervisor) that would have made that whole enterprise quicker to develop, safer (for OSX users), and more seemless if they had stuck with RISC-based PPC.
So my point, which really hasn’t been refuted, is that we Mac users could have had a great, seemless, performance upgrade path, along with all of the other bells & whistles that going to x86 provide, without a costly CPU transition. The latter has just slowed things down in many ways (computer prices going up, sales being off, & lack of truely new product introduction are my top three), and has given us nothing from Apple that we couldn’t have had before.
Apple’s stock price is a good indicator that more rational minds are seeing exactly the same circumstances. AAPL has stalled as a result of these things, pending proof that all this is in fact ‘a good thing’. I’m no advocate of focusing on stock prices to run a company, but often they give an unvarnished view of the consequences of actions companies have taken in real terms. In this case, the message seems to be that it is by no means clear that Apple needed this headache. And unless AMD really does implode (as all Mac fans will hope for as long as Apple stays tied to Intel), and/or something REALLY great comes out of Cupertino soon (something without major caveats of heat and battery-life issues, or integrated graphics failings, or feature ommissions – all apparently the outcome of cost containment needed for the more expensive Intel hardware) that can demostratably only have been possible with Intel, I predict investor conclusion will be that Apple’s spent a lot of money for nothing.
Apple did this for video. That market seems to be spiralling up, without Apple being at the forefront. I really want Apple to succeed, but if they never get the only benefits from going to Intel that they ever really counted on, that’s the sign of a bad management decision. A decision that we now know was made very suddenly, and NOT because Apple had no choice. And – as I said in my first post – if clear and obvious benefits don’t appear, after all this Sturm und Drang, then the buck stops at one desk – Steve Jobs’.
That’s my take & I’m stickin to it
” width=”19″ height=”19″ alt=”cool smile” style=”border:0;” />
Something that just occured to me as I read this article, and it’s seed, is that Apple almost certainly has a large hidden motive up their sleeve. One that they haven’t mentioned in all this discussion.
Apple is maintaining that all apps be made Universal. Why? Speculation, naturally, is so that they may switch back to PPC, as needed, to deliver the absolute best performance possible. Additionally, this gives them true hardware independence, from a microprocessor standpoint.
While this is logical, and likely true, there is something else to consider. Take a look at the largest software “partners” and ask yourself how many of them are still using Carbon? Adobe is, in, to my knowledge, every app they make for Apple. Carbon is completely outmode by this point, and Apple would naturally like to get away from it.
A hidden, undiscussed benefit of the Intel switch is a forced rewrite, for “performance”, of ALL code, to avoid the whole Rosetta performance issue. This kicks software vendors in the ass to get them to finally switch to Cocoa, and ends the reliance on now 5 year old workarounds to the upgrade to OS X. Once this process is complete, and all apps are Cocoa-based “Universal”, Apple is free to use ANY processor they like, taking full advantage of any architecture simply by tweaking the Cocoa libraries that ship with the OS and are automatically updated on every system.
This certainly means that IBM, Freescale, and even PA Semi are by no means out of the Apple race, if I am even remotely on target here.
Remember kids, you heard it here first.