Slashdot Mirror
Why Apple Picked Intel Over AMD
An Anonymous Reader writes "Macworld has a piece looking at why Apple chose Intel chips over AMD's offerings when it decided to move away from IBM." From the article: "The reason, industry analysts say, is that Jobs has a clear goal in mind: innovative designs. And such designs require the lowest-voltage chips, which IBM and Freescale were not going to make with the PowerPC chip core--and which AMD has not yet perfected 'This is a practical, pragmatic Steve Jobs decision,' says Shane Rau, Program Manager, PC Semiconductors for market research firm IDC. Intel serves up the most complete line of low-power chips for mobile and small form factor computers, and a good-looking future roadmap for it. Also, Intel's mammoth production capacity erases any supply worries. "
Maybe massive, cost-saving volume discounts were a factor too?
I know this is Slashdot, so you didn't rtfm, but how about reading even the summay: "Also, Intel's mammoth production capacity erases any supply worries."
Nobody else in the industry has the capacity to produce like Intel does. Transmeta least of all because they're a fabless house.
- ------- There are ten kinds of people in the world. Those who understand binary, and those who... Huh?
I think this is the first time I have seen the words "practical" and "pragmatic" in the same sentence with "Steve Jobs". Remember the reality distortion field?
My co-workers and I pretty much sussed that out the day of the announcement. Others have, quite a while ago as well.
Said that it is worth while to mention that IBM is not incompetent. Their embedded cores which are custom designed are even more energy efficient. But again they are expensive (and task specific) and cost drives the market.
Because apple doesn't care about top-end performance.
Have Apple PCs ever been ahead in performance? Of course I'm talking about real performance, and not ridiculously narrow, artificial tests to highlight a largely irrelevant strong point.
I don't mean this to discount Apple, and the truth is that virtually any PC (PC including Apple) these days is overpowered for the uses that the average user tasks them with, but I just don't buy the mythology of the hyper-super-mega PowerPC chips - always barnburners on paper when they're long in the future.
I mean, the decision has been made. They're going with Intel processors. At this point, I don't think it matters much why they chose to make their decision. Regardless of why they made the transition, we're all going to have to live with it. We'll have to port our software, and if we want new systems from them we'll just have to accept that they will have Intel processors inside them.
Perhaps there are better questions to be asking. Namely, what can we do with these new systems that we could not do before?
Cyric Zndovzny at your service.
It's all about the mobile processors. Intel's PentiumM's are FAST, low power usage and did I mention FAST? Seeing as how Laptop purchases are rising faster than desktop purchases, and since Apple's laptops are the most long in tooth, I'm betting that the first new Apple Intel box will be a laptop.
If AMD comes out with a better chip in terms of power usage, Apple can switch anytime. As such, going with Intel at the start implies no committment. If Intel starts treating them like dirt, they can go over to AMD, or even perhaps VIA. That's a choice they didn't have before with the PPC architecture.
Oh, I see he wants innovative. Since intel has been so innovative the past few years, it's easy to see why it was such a good choice.
Wait, wasn't it AMD that stepped up with the 1Ghz cpu first?
Oh, weren't they the ones who got the first high performance, low cost 64 bit processors to market?
Geez, haven't they also been dominating the performance side?
Besides, from what I've been reading, the Turion 64 is not far away from the Pentium M. Close enough to call them comparable at least, and the Turion has 64 bit extensions!
And they sell such huge quanitites that supply is a very important issue. Not.
Intel bunged them in the form of huge discounts, simple as that. No one in their right mind would use Intel processors for desktop machines at the moment and, for that matter, there's no reason Apple couldn't have gone with Intel for the laptops and AMD for the desk.
ALL of which is beside the point that the problem with the PowerPC seems to have been on the compiler side, not the hardware.
TWW
"Encyclopedia" is to "Wikipedia" what "Library" is to "Some people at a bus stop"
When will people ever learn. Ok now lets think for a minute. Apple can use Intel or AMD. Intel is able to guarantee more volume than AMD at the moment. This will change when AMDs new fab comes online this year. Intel also promised Apple a taste of the ol' MDF pie. MDF (Market Development Funds) as they like to call it helps Apple compete better with the likes of Dell in this space. MDF also guarantees that Apple will use only Intel CPUs. Now unless our heads have been buried under rocks for a while we all know that AMD technology is superior to Intel in sevral ways at this point. But Apple chose the inferior technology because Intel promised it massive $$$ kickbacks. Intel basically "buys" its customers. This is not rocket science folks. My prediction. After the move to x86 is stabilized Apple will then be free to use AMD as a tool to get better deals from Intel as Dell currently does. Due to their volumes they will not be able to get the prices that Dell gets so they will unlike Dell introduce a line of AMD cpus in the future. For two reasons. To have the MAC daddy of all X86 PCs and to stick it to Intel.
Yep right, no supply worries there, just ask any taiwanese OEM who wants to buy Centrino bundles ...
Did anybody else notice that Jobs in his keynote addressed why they're switching to Intel, and now however many weeks later the analysts put pen to paper and write down what he said as the reason they think they're switching?
Finance tutorials and more! Understandfinance
This has been dissected before, by an ex-Apple engineer. Apple went with Intel simply because of cost. AMD makes low-power CPUs, too (HE and EE Opterons), and they run circles around Intel's, but they cost more than twice as much to manufacture.
By using x86 CPUs, Apple has effectively lost the possibility to claim that their systems are magically faster than PCs (they never were, but they could claim it, because some people will fall for anything). So why pay more when the best they could aspire to was a claim that they were "on par with the fastest PCs"?
And, of course, there's another element: DRM. Intel cut Apple a good deal because it gives them a chance to start edging their hardware-based DRM into the market (think iTunes). Apple is happy to include DRM as long as they get a discount on the hardware.
A lot of ink has been spilled on why Apple chose Intel over AMD. I think it's all a bit of a waste of time.
Unless Apple uses some proprietary Intel instruction set, it can add AMD offerings to its lineup whenever it feels like.
My guess is that Apple chose Intel for their arch switch because:
1) It was easier to pick a single chip partner to do the switch with.
2) Intel likely offered incentives to go with them alone. There may be contracts involved in this, but they won't last forever.
3) Like it or not, Intel is the x86 brand with mindshare in the public eye.
4) AMD probably can't handle the volume of bringing all of Apple's products over to them at the moment.
The fact is that as soon as OS X is x86 it can benefit from the Intel/AMD competition in the same way that Windows and Linux users do today.
The hurdle is converting from PPC to x86. Going from Intel to AMD later on may not even be noticeable. In fact, if you think of the G4/G5 branding in the current Apple world, most consumers don't even know that their G4 is a Motorola chip and their G5 is IBM. They don't care, so long as there's an Apple on the side of the box.
Considering that AMD's chips are generally faster, cooler, and more efficient than Intel's chips, the choice of AMD would seem like a nobrainer without the Pentium M.
Apple seems to be moving hard toward mobile computing now anyway, so going for the Pentium M is a smart move all around, and it doesn't take much imagination to see those in Mac Mini's and the like in the future.
Myself, I'd have split the difference and gone with AMD for the 64 bit server chips. I think that descision is going to do good things for Sun.
ad logicam Claiming a proposition is false because it was presented as the conclusion of a fallacious argument.
LinosX? Let's see.
/. about Apple forking the kernel. Not to mention questions about GPL compliance and the bi-monthly call for Apple to opensource more of their stuff.
RSM would be running around shouting "It is GNU-LINUX-OS X!"
ESR would write a totally incomprehensible article claiming that "we won!". Again.
Linus would shrug and say 'whatever'.
There would be a flamewar on lkml between FOSS diehards and Apple engineers over their binary only drivers.
There would be frequent articles on
In short, it would be a lot worse. Linux would be pulled in two directions and Steve Jobs and Linus would be in a hornet's nest of unherdable cats trying to calm down tempers and get things done.
If J.K.R wrote Windows: Puteulanus fenestra mortalis!
Intel have a public roadmap into 2007, so their private roadmap must extend even further.
AMD have a public roadmap into 2006, but nothing long term. Privately, it may be different.
IBM have a roadmap into next week if you're lucky. Privately it may be different, but 3GHz G5s?
AMD has Intel beat at the moment on power consumption on the desktop, we all know that. However Yonah and Merom (and server variants thereof?) are what Apple are interested in. Yonah will come in many variants, with an ULV single core at 5.5W, and dual-core LV at 15W alongside the 35W dual-core standard processor. AMD have Turion however, and it isn't that bad in comparison with the current Pentium M, and 65nm should help them along even more.
It will be interesting to see how next year's processors compare. I think that AMD will remain leading in terms of performance at the high end, but the mobile arena will become very interesting with dual-cores from both company, new 65nm processors, and more to boot.
Sure, AMD holds the top-end. Not by an all that huge margin (say 20% on average to be generous) compared to how CPU wars have played out in the past. While AMD has gone from being the absolutely clear bang-for-the-buck manufacturer with the K7 to being the top-end holder with the K8 however Intel has really improved the rest of their product lines. A much overlooked chip today is the (new) Celeron D. 64-bit capable, solid performance, rock-bottom price. I would personally say that Intel offers better budget solutions at the moment.
Other than that however, I have said it before and I'll say it again; Intels desktop Pentium M roadmap can no doubt look great. The Pentium 4 did not work out as they wanted, but Intel has a lot of great engineers (just look how well the Pentium 4 has carried on competing despite the setbacks the design has seen), when they with the next big iteration are freed from the old P4 there will no doubt be a lot of interesting stuff coming from Intel.
Another interesting point is that Intel really is the only CPU maker that actually does more than one product-line at once. AMD kept the K7 around for a budget-line and stripped down the K8 a bit for laptops, but Intel has not just two, but actually three current designs ongoing (the P4, the Itanium and the Pentium M). An Intel roadmap may also contain a lot of goodies directly deriving from the fact that they have the design manpower to actually work on more than one path at once.
When Apple announced the switch, the roadmap has the transition beginning with "value" Macs and portables in mid-2006, with the rest of the line transitioning over the next year.
Basically, they will replace G4-based systems first (eMac, mini, portables), since the G4 Macs are currently the most clock-speed restrained. G4 processors are pretty low on power consumption but top out under 2GHz.
The G5 desktop roadmap is good enough to keep going for a while, with small clock speed improvements and a probable move to dual-core G5 chips. Apple also makes their highest profits on the G5 desktops, so they've got an incentive to push that as gently as possible. Look for the switch there to be right to dual-core x86-64 processors. Right now, G5 processors are still competitive with their x86 counterparts, so that's the other reason to concentrate on the G4 models first.
Hopefully they'll change Xserve last. Those things are pretty darned slick as-is.
Vista is currently due at the end of 2006 (about when Apple plans to release Mac OS X 10.5 "Leopard"), so Apple should be well into the transition by then. If Vista slips any further, Apple could even be most of the way through the whole process.
-- Josh Turiel
"2. Do not eat iPod Shuffle."
In the future. I'd imagine that Apple could fairly easily add AMD to the line where it fits (an Opteron in their server line perhaps). I can't believe that Apple has taken a completely "Intel only" approach on this...but I'm usually wrong on this sort of thing.
The cloners only made PowerPC-based Macs, not 680x0. Plus they had to buy their chipsets from Apple as well - they just tweaked them for higher performance than Apple was willing to do. Apple still sold the vast majority of MacOS systems (I think the total clone marketshare never exceeded 15% or so), but the thing that bit Apple about it was that that clone market (especially PowerComputing and Umax) was taking the highest-end part of the Mac market. And that was where the biggest profits were.
Jobs used the G3 transition and the accompanying move to "MacOS 8" (which was really just 7 with a few things bolted on, not the "Copland" 8 that was originally planned in the licensing deals) to freeze out clone licenses and get the market back. A nasty trick, but it worked.
I still remember Macworld Boston that year (1997), when everybody but Apple announced G3-based desktops shipping RSN (as soon as the licensing details were worked out with Apple). Of course, that never happened, so the tiny handful of those machines that ever made it into customer hands are probably collectors' items. Afterwards, Apple came out with their thoroughly underwhelming G3 desktop, and continued the death spiral...
Until the iMac was unveiled in 1998, and all of a sudden Apple started to get their mojo back. The rest is history.
-- Josh Turiel
"2. Do not eat iPod Shuffle."
Redundant?
No, try "true". If you know anything about Jobs and his past with Motorola, this becomes an very highly probable conclusion.
Jobs lost out because IBM doesn't need to put up with his shit now that the consoles are on the way. They weren't making IBM enough to be worth the necessary R&D to keep Jobs happy, so they told him to go fish.
The 64 bit extensions that Intel uses are a JOKE. They make the processor take turns when deciding what code to execute, while the AMD keeps chugging away and thus makes their 64 bit solution much better.
The Celeron D is okay, but compared to the Sempron again, a joke. You're talking about a chip that is just cheap (and relatively slow) to a Sempron that is cheap and relatively fast. Not to mention, it runs cooler too.
I think I'll echo what everybody else has said. It's a combination of money and laptop options.
Besides Apple can always switch to AMD at a later date.
The price is always right if someone else is paying.
That's misleading, because Vista doesn't really have any features, itself, except pretty pictures and a nice new browser. If you're talking about the forthcoming WinFS and Indigo, which ARE significant things, you're also wrong. Spotlight isn't really a database file system, but is more of a quick solution thrown together that looks pretty, but really isn't anything we haven't already seen on Windows in various forms for quite some time now. It's just got a nice menu icon. Apple tends to be good at getting things out quicker because they are content to wrap a pretty interface on a quick and dirty fix. (Which actually sums up Mac OS X itself, pretty much, which is a total kludge of OpenStep on BSD on HFS, and none of the three are completely integrated yet.) However, if it looks good enough to please the art school crowd, it works for Apple and everybody forgets that the "new" operating system is really a 15 year old class library.
So, the REAL question is: when Microsoft finally gets their shit together (they will eventually) will Apple have something to really compete against WinFS, Indigo and C# (which is lightyears ahead of OpenSTEP's frameworks and 'objective' C)? My hope is they do (I'm typing this on a PB G4) but my point is that it's not fair to say Tiger already has any of this. Microsoft, for all their faults, tends to try to do things the right way, at least in terms of computer science, if not morality, whereas Apple just tries to get it done. In the end Microsoft will have an almost new operating system with some very modern ideas on file systems and distributed computing. For all the delays, it could be quite good, actually. Apple, on the other hand, really hasn't added anything of huge significance to the old OpenSTEP except pretty pictures and a really nice graphics system. Yet...
They're analysts. They're smarter than us. Examples:
"I believe this is a purely negotiating move by Apple to grab some attention and headlines and to point out that they're feeling underappreciated by IBM" - Evin Krewell, editor in chief of the Microprocessor Report, quoted in the Mercury News, May 24, 2005, a few days before Apple announced a switch from IBM to Intel processors.
"You just wouldn't do that. You wouldn't do something that disruptive.'' - analyst Tim Bajarin, quoted in the Mercury News, May 24, 2005, a few days before Apple announced a switch from IBM to Intel processors.
"Stick a fork in 'em - this Apple is cooked." Robert Thomson, Financial Post, 2/20/2003
"For those who love Apple's products, this is all just so typical. This company has made an art of innovation -- from the personal computer itself to the point-and-click operating system -- only to invariably surrender the high sales ground to the boring knock-off artists who copy Apple's best ideas into a new and slightly cheaper model. So it's not surprising Wall Street is already bracing for another disappointment." - Steve Maich, Macleans.ca, 2005/05/09
Count David Goldstein, president of the Dallas-based growth-strategy consulting firm Channel Marketing Corp., among the critics of Apple's retail plans. "It makes absolutely no sense whatsoever for them to open retail stores," he says. - May 01, 2001 Macworld Magazine
I collect quotes like these, to remind myself why trusting analysts about anything is generally unwise.
Read the EFF's Fair Use FAQ
I don't care how many times I have to post this.
switching to Intel processor means you can run Windows software under OS X using Wine
this lowers the cost barrier for Windows users to switch.
Apple isn't just going to Intel for CPUs. Intel has all kinds of other chips and technologies, and at last they have a PC-making partner that will actually use cutting-edge stuff.
... CPUs.
And don't forget EFI. I doubt Apple's going to want a crufty old BIOS designed for 8086 machines. Intel has been working on superior alternatives to BIOS (although perhaps not as good as OpenFirmware, but still...).
AMD makes
He who lights his taper at mine, receives light without darkening me.
Intel's 3 lines (according to you): P4, Pentium M, and Itanium
AMD's 3 lines: Athlon64, Turion64, Opteron
Huh, go figure - both manufacture desktop, mobile, and enterprise/server chips
And come on now, the Celeron D? Mediocre performance aside, 64-bit Celeron D's start at about $80 street price (going off of newegg right now), but AMD's Sempron64s start at 50-60 bucks - I fail to see how Intel has the better budget solutions.
To echo what everybody else already knows - the Pentium M is a damn good chip, and it'll be the basis for the first Intel Macs, with Yonah down the line. That aside, Intel used no small amount of monetary influence in getting this deal.
Right. So instead of trying to get a chip manufacturer to bend to their product plans, they picked a manufacturer whose plans matched their own.
Apple doesn't want the latest fastest parts. They want the low power parts. Not everyone is a boutique consumer, only hardcore gamers care about the top of the line. Don't believe me? Just look at the volumes being sold. Boutiques is 1%.
https://www.accountkiller.com/removal-requested
It's about Dell - Intel's stalwart (yet irritating) partner. Dell has been playing 'footsie' with AMD for the last few years and in the process, getting more consessions from Intel (who is probably responsible for half of their advertising budget). Intel knows that Dell fears Apple (and make no mistake - they do).
This is really Intel's way of getting some of their leverage back. If Dell tries to pull one of those, "Well... You know... AMD is offering...", then Intel will be in the position to tell them to do what they like.
"...Well, there's egg and bacon; egg sausage and bacon; egg and spam; egg bacon and spam; egg bacon sausage and spam..."
WHAT?!? IBM is not going to make low-power chips?!? What about the 970 low-power line (13-16W) that EVERYBODY KNOWS?
Apple will be using PowerPC chips in at least some of it's machines through 2008. There is little reason to think that they won't use the 970FX (low power) or 970MP (dual-core) in a machine between now and then. I think the point you're missing then is that Apple saw what was comming from IBM and still decided to move to Intel. In other words, 13-16 watts is still too much.
According to Intel's presentations on the Conroe/Merom architecture, due 2H06, they're anticipating typical draw down to about 5 watts for the mobile version, and (IIRC) 25 watts for the desktop.
[Intel's] performance per watt numbers are the worst of the whole desktop industry.
And yet, their performace per watt numbers for mobile chips are the best. You seem to be implying that Intel for some reason can't design a low-power chip, when it's quite clear that they can.
And then intel promises apple CPUs which give 5x more "performance per watt". Yeah - that's nice when you consider that they get that "5x" number when they compare it with the current intel chips - which, as everybody knows, they're the worst at performance/watt.
What is your point here? Just because their current chips are the worst doesn't mean their chips next year can't be the best. Five times better than Intel's current might only translate to two or three times better than AMD's current, but that doesn't change the fact that the chip only draws 25 watts. Again, you seem to be implying that Intel can't possibly make a lower power chip just because their current chips draw a lot.
But heck, I absolutely hate how most of apple zealots just don't think - they repeat everything which Jobs tell them
What you should hate is people who assume that they can't possibly be wrong, and that anyone who disagrees with them must be incapable of drawing their own conclusions. In short, yourself.
Don't become a regular here -- you will become retarded.
It's simple...
Intel has vast software and development resources specifically to help assist in migration to it's processors from rivals. (Although this may be the biggest such case.)
Their resources in the software, compiler, etc. arenas is unparalled. AMD might be pumping out some great chip designs but I seriously doubt they could offer the transition resources of Intel.
However, once Apple is transitioned to x86 and their exclusive contract (5-10 yrs I would guess) with Intel expires. They will then be in place to take opportunity with whichever manufacturer has the better offer at the time.
So essentially, it was a wise long term strategy. Choose the one who can offer the easiest transition as in 2-4 yrs (after they fully transition) who knows who's chips will be faster/cooler/cheaper? After that time. If there is a better alternative chip it would be minimal work to allow for using an AMD x86 as opposed to an Intel x86.
Plain, simple, intelligent....
And Intel has better laptop processors. That's why SJ chose Intel as a primary supplier. When AMD gets its act together wrt laptop processors and kicks Intel's butt there, I won't be surprised to see AMD chips in Apple products side by side with Intel (unless Intel cut them a really hefty discount in return for exclusive contract).
With 4 billion dollars, 5 years and 575 project managers matrixed over a 700,000 step process. I'm sure that the day they succeeded they would all be downsized anyway.
See for all its tough talk about innovation, IBM and I suspect any other large command and control organization that's tried to outmanage and outprocess itself out of every dilemna by becoming even more bureaucratic really can't move quickly to do the right thing. And even when it succeeds at moving at all, it's typically the wrong solution poorly executed and overloaded with everyone's personal agenda items.
Moving to a company like Intel which for the most part makes chips and nothing but chips is usually the wise choice for a company looking to use chips. At best IBM's chip division, while capable and smart is only a division and one that gets the shaft more often than not because it's a supplier to all of the other IBM hardware units which are themselves victims of their own bureaucracies.
And if truth were told, if IBM thought there was money to be made in low power chips they would have done it already. Clearly IBM made a decision that Apple's goals did not fit with their own business model.
IBM foucused on all three console makers
AMD Solid high end chips but lacking in low power mobile area-small company
INTEL Focus on the computer market - dominating the portable market - big company ie greater resources - fallen behind at high end but starting to catch up
Add to this that previously Apple has always had to make there own chipsets, being able to use Intel off the shelf chipsets will make a huge difference in Apples margins.
plus you gain better access to Wi-Max, USB, Xscale
Nope hardware no longer makes the computers any different because it isn't different.
We are going to be stuck with x86 and it's 64bit mutants until the end of time.
Maybe it is because I remember when their was some difference in computers. Lets see which PIV with an Intel mother board and an ATI video card I should buy!
Or which AMD system with an NVidia video board runs Doom3 the fastest.
Boring.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
With Yonah and Dothan, Intel is already ahead on power in the category of "non top speed chipset". So I don't see how Apple is making a mistake here.
Apple is taking a risk that Intel will not be able to reform their top of the line chips to match AMDs superb offerings. But honestly, tower configurations don't account for much of Apple's sales anyway, so it's not a huge risk.
Anyway, I as I've said before, I think there are other reasons Apple chose Intel over AMD. To get Northbridges with integrated graphics and chipsets with integrated wireless, etc. is the main reason. Apple mostly makes all-in-one type machines. In order to be cost-effective, these machines will have to be much more integrated than they currently are. And AMD doesn't offer much besides a processor. They don't make chipsets right now.
So I don't think AMD really made much sense to Apple at this time.
http://lkml.org/lkml/2005/8/20/95
Another thing I would note is, how many of the companies on this list turn a consistent profit, who aren't named Dell or Apple. HP may sell more machines than Apple, but seem to loose many millions every quarter.
I have intel and amd computers and amd has a dramatically better cost/performance ratio.
yes, but apple was looking for performance/watt ratio. Apple wants to sell more portables and SFF PCs.