Apple set to go Intel? (Again)

[Penman]

You're saying this as if it's somehow new or relevant, even though it's been known that Apple has been keeping x86 operability since the days of the Rhapsody project. Apple wouldn't be dense enough not to at least test the basic OS on x86 hardware. The problem is in the fact that you'd lose quite a lot of your software when developers had to port. Again. You know, right after moving to OS X?

Ports to x86 are commonplace and there's very little software that exists that won't run on a x86 chip. It's not like moving to PPC.

Yes, we do, and that's why we're on PowerPC instead of the dying x86 spec. Intel's as good as admitted that the P4 is a dead-end line of speculation at this point and that they're going to have to move to the Dothan/Jonas/Merom line of processors on the desktop. AMD has long realized that clock isn't all there is to performance and has been keeping pace with Intel while sticking under the 3.0ghz mark.

Efficiency is key.

x86 is an old design but difficulties don't mean it's dying. Intels plans still allow for massive compatibility and even a change of core architecture (which is now happening) doesn't mean the end of Intel. Hating Intel doesn't make their product bad. They're reacting to problems extremely aggresively (dropping the P4 roadmap). IBM are struggling on. What seems like the smarter approach to you? Intel will do ANYTHING to maintain their competitive edge. That's not a bad train to be on.

As for the fantasy that the speed wars are over - it's laughable. Architecture's as important as clock rate but it's no coincidence that dual Opteron Systems at 2.4 GHz benchmark close to dual G5's at the same speed and G4's if they ran that fast. As they say with cars - there's no replacement for displacement. Clock speed is cubic inches. Jobs knew that when he promised 3GHz last summer. When he delivers on that will you be calling him misguided? I don't think so.

I hope Apple uses the best processors they can get, not the cheapest.

By who's measure?

If Apple sold Intel, they'd be under within two years. A huge portion of the usability of the Mac OS is that it's on a controlled platform with a limited ability for others to hack into the firmware and either alter or pirate the code. If they were to move to x86, the bootlegs would be out within a week and the hardware business would sink.

Apple's left alone by pirate and hackers because it's small. Not because it's secure or unworthy. Don't be fooled.

Apple is not a commodity company and PC users are not in the habit of paying top dollar for design.

That's blinkered and arrogant. Alienware and others show that design and quality matter to PC users. Just because Apple products put design first in many cases (and succeed spectacularly in doing so) doesn't mean that people who don't buy Apple don't care. The new BMW's look as if their constipated and I'd still buy an M6. I'd just know it's not too pretty.

It would be the best thing to happen to them in a long time. Apple on x86 hardware would be even easier to crush, since Microsoft wouldn't even have to develop VPC anymore. All they do is sell an application to tap the Windows APIs or ease dual-booting for those who aren't power users.

You can't out-Dell Dell, and you can't beat Microsoft in a price war when you're only the size of Apple. That's why Linux on PowerPC, with the weight of Apple, IBM, Toshiba, Sony, AMD, nVidia, Freescale, and others is going to be what it takes.

Linux is a joke. It's not ever going to be a desktop replacement because it's too geeky and doesn't improve on windows in a way that non-geeks care about. Manufacturers don't want what's best (and Apple are included here) they want what's best at making them money. Linux is not it. Microsoft have earned a place in computing equivalent to the utterly ridiculous QWERTY kwyboard and the mistakenly reversed numeric keypad. None of those things are going away. Apple will survive in good shape if it sells more computers. People who buy Macs are the least technically adept consumers on average (and I've seen the research). If you provide an Apple like experince via software and OS few care what's under the hood. Apple have abandoned SCSI and taken on PCI and USB without any ill effects, an Intel chip's just the same.

 

[alandail]

The point of my long post? A low cost Xserve (though Apple should probably create a new brand name to avoid confusion with their PPC based counterparts), (BTW, low cost spelled $500) based around a cheap X86 processor AND powered by a version of Mac OS X Server, built specifically APPLE based X86 processors (not sold separately for other X86 machines initially), which comes with all the tools and applications the current version of Mac OS X Server for PPC machines comes with (the built-in email system, DNS system, file server systems, etc.) would sell like hotcakes to small businesses (small like 1-10 employees) who do not have the need or resources for an IT department and who have cheap broadband access (our business has a DSL connection from SBC for under $40 a month, we could upgrade to a package for $60 a month that would give us more than enough bandwidth for an Email and Webserver AND still be able to download work related materials.

If Apple were to release a server with the above specs, for $500 dollars or below, that would basically allow me to setup an email and webserver in a matter of minutes, I would buy on in a second. I think you will find a lot of people like me out here.

What would the advantage be of using x86 instead of PowerPC? And doesn't an eMac pretty much meet the description you just gave? Not MacOS X server, but is MacOS X, which is unix based. It takes all of 5 seconds to be a web server under MacOS X. No way you see a MacOS X Server machine for under $500 - they charge $499 for MacOS X server itself.

 

[JFreak]

i think this is just a rumour, but there might be a partial truth here...

remember that amd chips, though they are x86 compatible, are in fact risc processors. amd chips only emulate x86 code and couldn't run windows before microsoft wrote drivers for them. so they are different beasts than intel's, and as i have understood, amd chips behave very well and are manufactured by ibm.

so there you might have the point of the rumour, as apple could make an amd-compatible version of osx leaving intel users unable to run it. while intel chips seem to be below apple standards, amd chips are competing the G5.

and at the end of the day, IT ALL COMES DOWN TO NUMBERS. let's say apple makes $500 profit from each computer sale they make (which is most likely not the case, but let's imagine) - and that includes both hardware and software. now take a look at the apple store and find out that the os costs $129 and take it off the profit, and it comes down to $371, which is about three times the os price. as we all know, apple is very profitable with current sales for the 3% and this loyal user base probably will buy apple hardware in the future no matter what. they might just calculate that just by growing their market share from three to ten dollars (by selling amd compatible os without hardware) makes them AT LEAST as much money that they will lose because some portion of the three percent would likely switch using cheaper amd hardware. they probably make even more profit by doing this, because if they don't let the company bloat as microsoft, the operating costs will not go up ---> all amd-compatible os sales are pure profit.

so that's just numbers, and they show that amd-compatible osx would be profitable.

but what about apple quality standards? will they be able to say "you can run osx on any amd compatible pc out there"? most probably they would go the same route digidesign is going (digidesign is the company that makes protools, which is the de-facto standard in recording studios all over the world), by specifying that these and these configurations they have tested and are 100% compatible, but other configurations might work also. they don't promise anything for configurations outside their specs, but many people have had success running protools with un-specified hardware.

so there you go, that's possible. not likely, but possible. it is possible apple is dividing itself internally as a software division and a hardware division. they both can be made profitable, and one of them can lose money (short-term) while keeping apple profitable as a whole. apple can very well decide that their hardware division would have to lose money - say, for five years - in order to software division to gain market share. when they have more os market share, there is also a good possibility that more people will switch to apple hardware.

but apple and intel? that will never happen. i can see someone putting a porsche engine on an old beetle, but not the other way around... (yes, i know, car analogies are stupid.)

 

[thatwendigo]

Low cost Servers using industry standard X86 running an industry standard UNIX operating system (BSD, the core of Mac OS X) WITH one of the most well developed GUI every devised, BUT with the compatibly of an Xfree86 server are exactly where Apple should focus their efforts.

...

If Apple were to release a server with the above specs, for $500 dollars or below, that would basically allow me to setup an email and webserver in a matter of minutes, I would buy on in a second. I think you will find a lot of people like me out here.

I'm sorry, but I have to lump this into the same place that I do the "iMac with removable screen and 3.0 billion ghz with top of the line graphics for $5000" category, not because you're stupid or anything like that, so please don't take it that way. You have a well-stated case and one that I can see that you've thought out a bit, but it doesn't take into account things like what I call the Dell-factor. No matter what Apple does with x86 hardware, someone will beat the hell out of them on the pricepoint, unless they break their tradition of service and quality in favor of trying to out-Dell Dell.

You can't get the GUI on x86, and I don't see that changing any time soon, for any market segment. The simple fact is that it would be far too much of a risk unless the OS was coded to a specific chip, and that would restrict the upgrade path later on. A $500 PC is also going to be a bit on the crappy side, unless they make no margin on it, and part of the whole point of OS X Server is to sell more PowerPC macs. That's why there's basically no client limit.

Ports to x86 are commonplace and there's very little software that exists that won't run on a x86 chip. It's not like moving to PPC.

It is if you're moving PowerPC-specific code across to the x86 architecture, which is what I was saying in the first place. Even for programs coded under Windows, you'd have to rewrite them for the OS X APIs. It's not as if there's magical x86 code that just runs on anything, though there would be less of an issue with porting certain kinds of *nix code than there would be for Carbon/Coca or WinAPI applications.

x86 is an old design but difficulties don't mean it's dying. Intels plans still allow for massive compatibility and even a change of core architecture (which is now happening) doesn't mean the end of Intel. Hating Intel doesn't make their product bad. They're reacting to problems extremely aggresively (dropping the P4 roadmap). IBM are struggling on. What seems like the smarter approach to you? Intel will do ANYTHING to maintain their competitive edge. That's not a bad train to be on.

First of all, make no assumptions about my feelings on Intel. I respect their design on the Pentium-M core, even if I don't choose to use their chips because I want to use OS X on an efficient desktop processor and all of my machines are pre-Centrino. That being said, I think that IBM is the smarter investment, because they have more money, more patents, and more IP than Intel could ever dream of. Big Blue files the most patent claims a year, owns the most patents in the world, and they own the processes everyone else is moving to. You've heard of SOI and SSOI, I'd assume, and those are both IBM technologies. While Intel and AMD would never call them such, their designs are more and more RISC-oriented, with a thin x86 compatibility layer on top. That's why the Athlon line has been consistently competitive or beating the Pentium line at lower clockspeeds.

As I said, Efficiency.

As for the fantasy that the speed wars are over - it's laughable. Architecture's as important as clock rate but it's no coincidence that dual Opteron Systems at 2.4 GHz benchmark close to dual G5's at the same speed and G4's if they ran that fast. As they say with cars - there's no replacement for displacement. Clock speed is cubic inches. Jobs knew that when he promised 3GHz last summer. When he delivers on that will you be calling him misguided? I don't think so.

I suppose it's easier to put words in my mouth than do a little looking at my past posting record, but what you'd find is that I have a pretty decent understanding of chip design for someone who doesn't work in the industry. Part of that is because I read actual, technical descriptions of how these chips works, usually from Ars Technica, but sometimes from digging them up myself.

I've freely admitted that the G5 is competitive in most places, dominant in some, and lagging in others. It's the reality of modern microprocessors, especially witha design philosophy that's been historically different from the competition. I want the next chips to scale, but I want them to do it efficiently. We don't need the 970 to become the Prescott of the newer core designs.

Apple's left alone by pirate and hackers because it's small. Not because it's secure or unworthy. Don't be fooled.

You misunderstand my point, then. Apple moving to Intel would make it so that there was no reason f or hackers not to just crack it wide open and put it on cheap boxes. Apple's sales would plummet.

That's blinkered and arrogant. Alienware and others show that design and quality matter to PC users. Just because Apple products put design first in many cases (and succeed spectacularly in doing so) doesn't mean that people who don't buy Apple don't care. The new BMW's look as if their constipated and I'd still buy an M6. I'd just know it's not too pretty.

Who sells more computers, Dell and HP/Compaq or Alienware? Who sells more Dell or Sony's Vaio? Who sells more, IBM or Alienware?

What's that?

Not Alienware or the other niche players?

Then it looks like most PC sales are still going to ugly, black-box (since beige seems to be out) units.

Linux is a joke. It's not ever going to be a desktop replacement because it's too geeky and doesn't improve on windows in a way that non-geeks care about.

Here's a dynamic for you, since you seem to at least know some of the market:
This past quarter, Sony, Toshiba, and AMD all bought PowerPC licenses and/or design rights from IBM. IBM is fabbing for nVidia, among others. The HyperTransport Consortium is huge. Take the different pieces, and apply it intelligently. IBM or Freescale fab a PowerPC chip to go on an nVidia chipset motherboard with embedded graphics (but an AGP slot) manufactured by a HyperTransport partner like Asus or Acer. This board goes in a cheap case, with a Toshiba hard drive and a Sony optical drive, providing basic functionality and needs for consumer use (USB 2.0, IEEE 1394, GigE, audio in/out). Commodity RAM goes in there, and the kicker... A well-developed, stable kernel drives the whole thing, just waiting for a designer to throw a GUI on it. Darwin.

A recipe for Linux on PPC, as sold to Enterprise and consumers, for the price of a budget PC and a slice for every unit sold to all the companies involved. It doesn't make a whole lot of money, but it makes enough to keep them interested.

Apple have abandoned SCSI and taken on PCI and USB without any ill effects, an Intel chip's just the same.

You don't have to rewrite the whole OS for a different processor if you remove SCSI support or add a new peripheral architecture. That's just a ludicrous comparison.

 

[rt_brained]

Hopefully this will happen one day. Maybe apples will be priced at what they should be if they go to x86.

Besides the typical arguments for or against, try and look at it even deeper.

If the rumor holds true (and I hope it doesn't) then I suppose Apple would be in a position once again to license their OS to companies like Sony, IBM, Gateway and others. If true, Apple will officially become a software-based company. Even if Apple tried to control licensing of their OS, there would be scores of people figuring out ways to make it work on scores of generic x86 machines anyway.

If they license their OS then there would be no point in building Apple branded machines anymore, as we all learned from the early 90's competititors' machines would severly undercut Apple's profits...The first thing Jobs did when he came back to Apple was put an end to OS licensing. Going back on that decision would create a huge PR problem for him because it'll make him look indecisive. Vasilating back and forth could hurt his credibility as a leader and a visionary. For all the flack he gets for his wild ideas, even his enemies secretly wish they were more like him.

Apple's visibility outside of it's customer base relies heavily on R&D and their slick innovative hardware (read iMac, Powerbook, iBook, iPod, Firewire, etc.). If that essential element goes away, then Apple has only their innovative software to rely on to draw consumers...yet many (most) Wintel users know little if anything about Apple's software or how it works. This would be brand new marketing territory for Apple to have to navigate. Are willing to accept that Apple is willing to throw its brand image out the window and start all over again?

The only specs of creedence I can give to the possibility of this rumor are:

1) IBM is not keeping up with their end of the bargain on their chip. Jobs promised 3 GHz by summer, but it looks more and more like IBM will miss that deadline.

2) Perhaps Jobs sees potential competition on the horizon from Red Hat and wants to cut them off at the pass to gain a lock on the Unix OS option. It's a long shot, but Red Hat has a lot of fans in the IT world, and continues to gain popularity among x86 users who've grown to hate Windows as much as the rest of us.

 

[thatwendigo]

If the rumor holds true (and I hope it doesn't) then I suppose Apple would be in a position once again to license their OS to companies like Sony, IBM, Gateway and others. If true, Apple will officially become a software-based company. Even if Apple tried to control licensing of their OS, there would be scores of people figuring out ways to make it work on scores of generic x86 machines anyway.

If they license their OS then there would be no point in building Apple branded machines anymore, as we all learned from the early 90's competititors' machines would severly undercut Apple's profits...The first thing Jobs did when he came back to Apple was put an end to OS licensing. Going back on that decision would create a huge PR problem for him because it'll make him look indecisive. Vasilating back and forth could hurt his credibility as a leader and a visionary. For all the flack he gets for his wild ideas, even his enemies secretly wish they were more like him.

Thank you for stating, a little more eloquently, what I've tried to explain about the move to x86 chips. It doesn't even have to be Intel in particular for this to screw us all.

1) IBM is not keeping up with their end of the bargain on their chip. Jobs promised 3 GHz by summer, but it looks more and more like IBM will miss that deadline.

They said summer, and it isn's even the beginning of summer yet. I love how people are naysaying this when the deadline hasn't even passed. Yes, there has been a bad history with missed ship dates. However... Who just canceled two entire processor lines?

Let's keep a sense of scale here.

2) Perhaps Jobs sees potential competition on the horizon from Red Hat and wants to cut them off at the pass to gain a lock on the Unix OS option. It's a long shot, but Red Hat has a lot of fans in the IT world, and continues to gain popularity among x86 users who've grown to hate Windows as much as the rest of us.

I don't buy it. Unless Apple stuck a really, really complex boot ROM on the boards, one that would be nearly impossible to crack, they'd still be shooting themselves in the foot. A bootleg copy of the business file would get out and they'd lose sales to people putting it on cheap x86 boxes and not even bothering to buy a copy. That's a complete loss for apple, instead of just a hardware or software loss.

 

[ftaok]

They said summer, and it isn's even the beginning of summer yet. I love how people are naysaying this when the deadline hasn't even passed. Yes, there has been a bad history with missed ship dates. However... Who just canceled two entire processor lines?

Let's keep a sense of scale here.

Well, since the jump from 2 ghz to 3 ghz is huge. The general consensus is/was that there would be at least one update in between 2 and 3. Since it's been 9 months and we're still at 2 ghz, it's looking more and more unlikely that we'll get 3 ghz by the end of summer.

Add in the reports that IBM is having trouble with the .09 micron process and Apple mysteriously missing 2.3 ghz X-serves, and you get doubts to IBM's ability to provide 3 ghz chips to meet Steve's promise.

The burning question is why can't IBM get the .13 micron 970 chip to go any faster than 2 ghz? It would have been nice to see a Dual 2.4ghz G5 with 970 chips. Has IBM abandoned the .13 micron 970 chip?

 

[thatwendigo]

Well, since the jump from 2 ghz to 3 ghz is huge. The general consensus is/was that there would be at least one update in between 2 and 3. Since it's been 9 months and we're still at 2 ghz, it's looking more and more unlikely that we'll get 3 ghz by the end of summer.

Oh, the general consensus... And would that hold any water if you took it to IBM and told them that a bunch of people got together on a website and thought they were going too slow? Nothing was ever promised to us, aside from 3ghz this summer. No incremental updates, no hardware revisions, nothing. N-O-T-H-I-N-G.

To blame Apple and IBM for getting your hopes up over rumors is neither fair nor rational.

Add in the reports that IBM is having trouble with the .09 micron process and Apple mysteriously missing 2.3 ghz X-serves, and you get doubts to IBM's ability to provide 3 ghz chips to meet Steve's promise.

Unless, as I've said, the 970 is a holdover and was never meant to be a flagship chip for Apple. I would not be at all surprised to see the 975/980 roll out at WWDC and be sold under the G5 flag, just like the G4e was sold under the same name. If you do the scaling math, the Power5 reaches 3.0ghz handily and the 970 does not.

The burning question is why can't IBM get the .13 micron 970 chip to go any faster than 2 ghz? It would have been nice to see a Dual 2.4ghz G5 with 970 chips. Has IBM abandoned the .13 micron 970 chip?

It's called heat. A 130nm 970 chip put out 50-55w at 2.0ghz, and so two of them is like one Pentium 4 Prescott. Ouch. I don't think you want to try to ramp that in a dual system.

 

[oingoboingo]

Oh, the general consensus... And would that hold any water if you took it to IBM and told them that a bunch of people got together on a website and thought they were going too slow? Nothing was ever promised to us, aside from 3ghz this summer. No incremental updates, no hardware revisions, nothing. N-O-T-H-I-N-G.

To blame Apple and IBM for getting your hopes up over rumors is neither fair nor rational.

That might not be entirely true if you were twiddling your thumbs waiting for an XServe G5 to be delivered. There have been delays past promised shipping dates due to IBM's slow PPC970FX production scale up.

 

[mr_mac]

Why go from the best Chip maker in the world to the worst??

 

[FFTT]

OSX on other systems

I posted my thoughts on this in the G5 update thread, so please forgive


IF Apple maintained a position where only the best professional systems
( Apple, IBM & AMD) were licensed to run on OSX and they made use of available "hard line" copy protection, this expanded market might convince
companies like ATI to give OSX compatible systems the priority it deserves.

Microsoft and Intel can continue to enjoy their marriage on cloned systems running Microsoft Longhorn.

baaaaaah

 

[ingenious]

Apple's left alone by pirate and hackers because it's small. Not because it's secure or unworthy. Don't be fooled.

we're not fooled. mac os x is much more stable than windows. its much more secure as well.

 

[Penman]

No matter what Apple does with x86 hardware, someone will beat the hell out of them on the pricepoint, unless they break their tradition of service and quality in favor of trying to out-Dell Dell.

That assumes Apple's offering the best service possible at its pricepoint. I like my Mac but don't believe it's great value. The service is good. The quality is good. I've no personal experience of it being outstanding - take a look at the lid of almost any Al powerbook (and I've bought 3) for evidence of mediocre design (the latch) and implimnetation (the build)

You can't get the GUI on x86, and I don't see that changing any time soon, for any market segment. The simple fact is that it would be far too much of a risk unless the OS was coded to a specific chip, and that would restrict the upgrade path later on.

Restricted to Intel is really almost no restriction at all.

It is if you're moving PowerPC-specific code across to the x86 architecture, which is what I was saying in the first place. Even for programs coded under Windows, you'd have to rewrite them for the OS X APIs. It's not as if there's magical x86 code that just runs on anything, though there would be less of an issue with porting certain kinds of *nix code than there would be for Carbon/Coca or WinAPI applications.

Most OS X software is seperately developed as is. The coders might complain but the concept is the same - once for Windows, once for OSX. For developers who accept that it's not too important that the new code wouldn't draw on the legacy code (though of course the sourcecode could be used verbaitim).

First of all, make no assumptions about my feelings on Intel. I respect their design on the Pentium-M core, even if I don't choose to use their chips because I want to use OS X on an efficient desktop processor and all of my machines are pre-Centrino. That being said, I think that IBM is the smarter investment, because they have more money, more patents, and more IP than Intel could ever dream of. Big Blue files the most patent claims a year, owns the most patents in the world, and they own the processes everyone else is moving to. You've heard of SOI and SSOI, I'd assume, and those are both IBM technologies. While Intel and AMD would never call them such, their designs are more and more RISC-oriented, with a thin x86 compatibility layer on top. That's why the Athlon line has been consistently competitive or beating the Pentium line at lower clockspeeds.

As I said, Efficiency.

IBM is a bigger company than Intel that files more patents. However IBM works in many areas of business and Intel is a specialist. Let's compare chip divisions and you'll see that any benifit is in Intel's favor, not IBM's. Despite IBM's patents it still sells less chips than Intel. I love RISC - I'm a Brit - we invented it at Cambridge in the 80's (you're welcome). However for all the CISC is dead arguments Intel still competes and wins.

Athlon is an exercise in reverse engineering. It's not even a secret. Athlon runs fast because they benifit from being able to stand on Intel's shoulders. The chips do some great things but they arrive only when the AMD engineers have had a chance to study mistakes and weaknesses in Intel's design. It's a lot easier to follow than lead (hence Intel's recent reverse engineering of AMD's 64 bit instruction set).

Efficiency is the argument Mac users use when they're behind. 2 2.0 Gig G5's equal 1 3.4 Gig P4 is hardly a paragon of efficiency.

I suppose it's easier to put words in my mouth than do a little looking at my past posting record, but what you'd find is that I have a pretty decent understanding of chip design for someone who doesn't work in the industry. Part of that is because I read actual, technical descriptions of how these chips works, usually from Ars Technica, but sometimes from digging them up myself.

I've freely admitted that the G5 is competitive in most places, dominant in some, and lagging in others. It's the reality of modern microprocessors, especially witha design philosophy that's been historically different from the competition. I want the next chips to scale, but I want them to do it efficiently. We don't need the 970 to become the Prescott of the newer core designs.

It should be understood that the 970 wasn't built for Mac. It was a server chip Apple decided on. To treat it as a custom design perfectly honed to the Mac platform is ludicrous. If that were the case it wouldn't run so hot and be entirely incompatible with notebooks in its present form. Apple are smart enough (I hope) to realize that G5 is just marketing. We could call any chip G6 and the public wouldn't care. It's just a brand. Intel's Itanium looks to be a very serious contender and Apple realize that IBM are never going to produce a chip whose sole aim is to compete with Intel's line - they can't win right now. The only people who want IBM to go head to head with Intel are the Mac faithful. It's not enough. Apple can choose to make do with server chips (and hope to sell enough computers to influence the design in a serious way) or move to chips designed for desktop computing. Apple sells an interface - as long as it looks and feels right the engineering's secondary. The engine for the new Mini is designed and built in the great automotive engineering country of Mexico - who cares?

You misunderstand my point, then. Apple moving to Intel would make it so that there was no reason f or hackers not to just crack it wide open and put it on cheap boxes. Apple's sales would plummet.

That wouldn't happen. How many 'pirate' DVD's do you own? Even if it could be cracked in software efficiently (which is highly unlikely given the other custom hardware - it'd run slow as molasses) - Mac buyers want the box and the logo. iPods sell on looks mainly. Most people don't realise that there are cheaper options which do exactly the same thing with better battery life and storage. If they do they don't care because they want an iPod. Macs are the same. It's an emotional decision (which is why people waste so much time here trying to justify their passion for the platform in the face of evidence it's not unquestionably superior).

Who sells more computers, Dell and HP/Compaq or Alienware? Who sells more Dell or Sony's Vaio? Who sells more, IBM or Alienware?

What's that?

Not Alienware or the other niche players?

Then it looks like most PC sales are still going to ugly, black-box (since beige seems to be out) units.

That's not a valid argument. Cheap outsells expensive. iMac outsells PowerMac. Is it because people don't value or desire PowerMacs? The reason ugly boxes sell the most (and you have to admit that style is subjective - I think Vaio's (and I have 2) are nasty purple things. I also thing that iMacs are kitsch.)) is because they're cheaper. If you give a PC owner the choice between ugly and pretty they take pretty. Saying that it's not true is akin to saying Ford outsell Ferrari because people don't value the beauty of Italian sportscars.

Here's a dynamic for you, since you seem to at least know some of the market:
This past quarter, Sony, Toshiba, and AMD all bought PowerPC licenses and/or design rights from IBM. IBM is fabbing for nVidia, among others. The HyperTransport Consortium is huge. Take the different pieces, and apply it intelligently. IBM or Freescale fab a PowerPC chip to go on an nVidia chipset motherboard with embedded graphics (but an AGP slot) manufactured by a HyperTransport partner like Asus or Acer. This board goes in a cheap case, with a Toshiba hard drive and a Sony optical drive, providing basic functionality and needs for consumer use (USB 2.0, IEEE 1394, GigE, audio in/out). Commodity RAM goes in there, and the kicker... A well-developed, stable kernel drives the whole thing, just waiting for a designer to throw a GUI on it. Darwin.

A recipe for Linux on PPC, as sold to Enterprise and consumers, for the price of a budget PC and a slice for every unit sold to all the companies involved. It doesn't make a whole lot of money, but it makes enough to keep them interested.

Linux has no marketing and no brand that anyone desires. Linux will end up in the home. It'll run DVR's, fridges and other embeds with custom front-ends. The desktop will remain with Microsoft or others who can offer the software people recognize and marketing to draw their attention. None of the licencees you mention would risk telling their consumers that the worlds dominant platform's no longer for sale.

Even if they did move to Linux you think they'd give it away and not extract a profit from consumers used to paying for software? Dream on. Apple repackage open-source code for their OS and charge us $130 a year (I think it's fair enough but am under no illusions that we're being profited from)

 

[Penman]

we're not fooled. mac os x is much more stable than windows. its much more secure as well.

I appreciate the enthusiasm and you are right on one thing. It's more stable (for very obvious reasons). However have you read about the serious holes in OS X that people have been discussing in the news recently. You could have driven a truck through them and Apple have been...ahem...economical with the truth when it comes to patching the holes.

 

[alandail]

iPods sell on looks mainly. Most people don't realise that there are cheaper options which do exactly the same thing with better battery life and storage. If they do they don't care because they want an iPod. Macs are the same. It's an emotional decision (which is why people waste so much time here trying to justify their passion for the platform in the face of evidence it's not unquestionably superior).

nobody else has the scroll wheel - How are you going to scroll through 10,000 songs with that stupid scroll button on the DELL?

 

[ThomasJefferson]

Bigfoot, Loch Ness, Black Helicopters, Area 51, Vincent Foster, West Palm Beach goes Buchanan 🙄 , .... Mac goes X86.

What would be the next progression of this illness?

Jobs to copy Dell's cutting edge style with interchangable Panther/Tiger palmrests?

 

[FFTT]

I'm sure there are a few pissed off people at Apple who are getting laid off
because something is up.

It frustrates me, more than anything to see how Apple tends to waste
a potentially easy opportunity to kick Microsoft while it's down.

Then again, we may not know what deals go on behind closed doors.