“I’ve considered–and rejected–this idea in the past. My feeling has been that, unless Apple were ready to cut the cord with Microsoft, it wouldn’t attempt this kind of head-to-head platform competition. Well, guess what? With Apple ads encouraging Windows users to switch platforms, and Microsoft whining about supposedly slow sales of Office for OS X, it’s clear that the relationship is in trouble. As the two drift apart, Apple has little reason not to make a processor change.” AnchorDesk’s David Coursey on the Mac/Intel issue.
Macs with Intel Inside? You Bet! Here’s How
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
If Apple decides to do this, then we will start to hear buzz from developers that are cross-compiling their apps to work with the new Intel version. It’s not something they could keep quiet for very long.
Another thought is that they could just come out with the XServer in Intel, as its used mostly for commodity apps anyway. Perhaps someone is running FileMaker on these things, I dunno.
But, unlike an LCD iMac, this is something that won’t be a surprise when it’s actually released, because a Mac that only runs Explorer and AppleWorks isn’t going to be very useful in the real world.
Also, if they decide to do this, then I’ll wager that there will be some kind of CPU key serial ID number thing in the Apple hardware to prevent the basic loading of OS X on PCs. (Which will immediately be hacked out of OS X, published on the web, and then slammed down by Apple using the DMCA…)
This dribble gets more funny by the day!
Is David Coursey in need for more hits?
> Is David Coursey in need for more hits?
I don’t think so. His article agrees a lot with mine. I like David… 😉
DVD ripping happens on every desktop, not careing about DMCA… mp3s are traded, software is traded… Apple will not slam anything once it has hit the wild…
Something’s gotta change at Apple very soon…powering their computers with a G4 is like putting a 4-cylinder engine in a Cadilac…OS 9 could get away with a slow CPU b/c it did less under the hood and was better optimized to the hardware…but OS X is huge, and poorly optimized as of yet!
personally, i’d be very surprised to see them go the way of x86 b/c they’d lose a lot of control…and control is really what Jobs is all about…i’d sooner expect Apple to go with a Sparc or have IBM make their PowerPCs
either way, i’d rather have easily upgradeable open hardware
just my 2c
-bytes256
I don’t think this will happen.
First of all, see how Apple’s current marketing strategy draw focus away from CPU speed to user experience. They don’t give you any tech specs, numbers or anything; It’s just “you can do photos, movies, music, email and web on it”, and the current G4s are sufficient.
Second: Apple will not be able to tie OS X to their computers with Darwin being open source. They already failed on the PPC: Where OS 9 doesn’t run on Mac clones, OS X does due to some Darwin modifications. Now, whatever proprietary firmware Apple may stick in their ficional x86 Mac: Darwin gives you a way to work around that. Darwin is already running on 440BX chipset computers, so programs running on top of that (ficitonal OS X Finder, Aqua, Quartz etc) will too.
come on guys!
They will be good PeeCees if they move to intellish CPU.
They need to be unique, what they need is to have good price performance ratio.
Maybe if they move to AMD Hammer .. they still need to find a way to say “this is a mac not a PeeCee”.
Apple going against Dell? Compaq? HP ? it will be a suicide IHMO.
If they would sell boxes able to run windows too … eh ! crazy
BTW .. all the optimization work done for years by developers on PowerPC ? lost? two adobe photoshop for MAc OSX? one x86-64-maybe and one PowerPC ? how adobe and others could afford such development efforts?
Maybe another emulation framework? running PPC code on Hammer? how many years to get it working properly?
It’s really an hard job today to be an apple developer I think.
If they can find a good technocal solution ok .. but .. maybe if I want an Hammer based computer I will go for a PC clone … why go for apple?
If Apple can find a way to keep its developers from having a cow at the thought of recompiling for another chipset, this could be very, very cool.
Someone (can’t remember where right at the moment, sorry) mentioned yet another interesting, although possibly improbable, alternative. Apple and NVidia seem to be getting pretty cozy lately; Apple has Jaguar, which is optimized for graphics card acceleration; NVidia has been getting into the motherboard market; what if Apple were to take (an admittedly huge) leap and transfer OS X to a completely new platform created by NVidia, where all OS functions are handled by proprietary, faster-than-spit hardware?
I know, I know … not likely. Probably all sorts of good reasons not to do this. But … what if they did it for their ultra-high-end line, aimed squarely at the graphics/audio markets? Could be interesting …
i still think mac on transmeta is the way to go
osx on oqo !
>>If they can find a good technocal solution ok .. but .. maybe if I want an Hammer based computer I will go for a PC clone … why go for apple?<<
It’s only PC/Mac users with common sense know that Apple going Intel would be the death of Apple!
1 mghz duals are still in the channel) is the date I hear that will render Courseys article moot.
“powering their computers with a G4 is like putting a 4-cylinder engine in a Cadilac…”
Ummm, actually there were 4-cylinder Cadillacs in the mid-80’s. Remeber the Cadillac Cimarron? It was a rebadged Cavalier.
Anything’s possible
Intel will have that much more of the overall CPU market. It’s sad.
If this happens … you can paint me X!
what if Apple were to take (an admittedly huge) leap and transfer OS X to a completely new platform created by NVidia, where all OS functions are handled by proprietary, faster-than-spit hardware?
Making a motherboard does not require making a whole new CPU Architecture. The NVidia motherboards take run-of-the-mill x86 processors just like all the other PC motherboards out there. I ahve yet to hear anyone praise the nForce as the bringer of speed, for any task.
I dont like the OSX/x86 talk, it scares me. I like Apple, i like OSX. Everytime i see the new iMacs in stores i keep thinking “maybe my next computer should be a mac.” They have an appeal to them. I think a lot of that appeal would be gone if they’d just to x86, it wouldnt be exotic anymore, it would just be yet another operating system for the x86. Isnt SPARC an open processor? couldnt they make their own and make more money with those?
“It has 2 1800 or 1600MHz G4’s, with 512k on chip L2, and 4MB L3 – a double pumped 333MHz , 266MHz ( it supports both, plus
the current 133, and other slow things )….ATA-100 RAID support –no start up penalty even with 2
GB DDR DRAM – Dual 80GB drives make a nice RAID .”
Surely Apple could do than Mac OS X only execute in their x86-box, but surely Windows will run over this x86 box and a lot of user could decide go to Windows. Are their users so loyal?
Second: Apple will not be able to tie OS X to their computers with Darwin being open source. They already failed on the PPC: Where OS 9 doesn’t run on Mac clones, OS X does due to some Darwin modifications. Now, whatever proprietary firmware Apple may stick in their ficional x86 Mac: Darwin gives you a way to work around that. Darwin is already running on 440BX chipset computers, so programs running on top of that (ficitonal OS X Finder, Aqua, Quartz etc) will too.
First off, I think if Apple were to build systems with x86 processors, they definately wouldn’t be PC compatible. I mean, seriously, a PC is just a souped up XT.
I would assume Apple would build a system without an 8042 controller and therefore no A20 gate. Having said that, they could build a system without an ISA bus entirely, completely free of all AT/XT legacy except for x86 itself.
I see this as leaving them exactly where they are now. You certainly couldn’t run Windows on such a system, although Linux and ilk could be modified to run on such a system.
On the flip side of things, the lack of XNU hardware support will work to their advantage. Even if you do manage to bypass option ROM checking, Darwin probably won’t run on your system because it doesn’t support your hardware, simple as that.
Certainly people can write XNU drivers for things like, say, network cards, but what happens when you want to use Quartz with something other than an Apple provided video card?
This article seems vaugely familer…
it would just be yet another operating system for the x86. Isnt SPARC an open processor?
I dissagree. They wont be selling box copies of OSX that wil l run on any x86 machine, you’ll still have to buy their ‘exotic’ computers.
It has 2 1800 or 1600MHz G4’s
Uh, no. I seriously doubt we will see 1600mhz G4’s anytime this decade. Much less anytime soon. And if your talking about AMD, they would not call their processors G4s for various legal reasons.
I used to want it because I was an apple fan who was whoo’d to the dark side by cheaper power, but now I want it to see all the poeple come out with “I told you so articles.”
” I seriously doubt we will see 1600mhz G4’s anytime this decade. Much less anytime soon. And if your talking about AMD, they would not call their processors G4s for various legal reasons”
Ah, no, I know what I wrote.
Not this tired topic again. Apple will not be going to x86. No way, for too many reasons. A more plausable scenario is going with all IBM PPC chips, which the company actively develops and pushes (not like Moto).
One thing I forgot to mention is that I’d miss PPC. Except for gaming, I can do pretty much everything on my 600MHz G3 iBook that my friends can with their 700-1GHz Celeron/PIII laptops. The difference is the iBook doesn’t have a fan, doesn’t make heat, and has much longer battery life.
Then I come home to my dual 1.53GHz Athlon monster, which generates inordinate amounts of heat, and wonder if the extra speed is really worth the cost you pay environmentally.
didn’t Apple snub IBMs PPC chips in favor of some Motorola ones in the past?
how funny would it be for a total reversal…it’s too bad really…Motorola had a pretty lucrative market in chip manufacture but through their f*ck-ups they’ve lost almost all of it…hell, even Palm is going intel (StrongARM chips)…another case of good technology and poor management
remember when Motorola powered everything from Sun UNIX workstations (the first ones were m68k’s until Sun started making Sparc’s) down to the tiniest embedded devices…hell they owned the world…but now they only wanna do cell-phones…idiotic bastards!!!
-bytes256
Certainly people can write XNU drivers for things like, say, network cards, but what happens when you want to use Quartz with something other than an Apple provided video card?
Write a driver – it’s as simple as that.
Like any other modern operating system, OS X is abstracting the hardware and nothing of the upper layers is communicating with the hardware directly. All the hardware abstraction is done in Darwin.
> Surely Apple could do than Mac OS X only execute in their
> x86-box, but surely Windows will run over this x86 box
> and a lot of user could decide go to Windows. Are their
> users so loyal?
For over a decade most Mac users have run virtual PC which lets them run Windows apps. Reminding Mac users of what Windows is like is how Apple keeps their users loyal.
I’m not saying that I disagree with the conclusion of the article (I’m not sure), but my lord what a poor article this was. It didn’t address any of the effects of such a move.
* All software would have to be, at the very least, recompiled. And now there are suddenly two versions to support. This would annoy the heck out of developers, and absolutely piss off users who suddenly would have to upgrade everything again, especially after all the recent upgrading to accomodate OS X. Even if the move to x86 was permanent (killing off PPC lines over time), the next few years would be an amazing headache, not only for Apple, but also for its developers and users.
* They would risk decimating their brand, which they’ve established over 18 years. The brand is worth billions to them. The whole point of the Mac is that it’s supposed to be a friendly little world where everything works together. They would go from having “Mac” to “Mac – PPC” and “Mac – x86”. That’s a big difference. Suddenly “Mac” software only works with some Macs. Ditto for hardware. Even with friendly names this would be disastrous; the brand is shattered. Imagine if you had to buy a different CD depending on the CD player you owned. BeOS fans were all geeks who have no problem with this kind of distinction, so they thought it was a non-issue. Believe me, it’s important. Apple is essentially a one-brand company; damaging that brand has the potential to kill the company.
The rather painless move to PPC from 040 was possible because they not only provided remarkable emulation, but also because the PPC was so much faster than the 040. Fat binaries allowed PPC users to feel the speed without leaving previous customers in the dust. No doubt that x86 is faster than the G4s right now, but not to the degree of 040–>PPC. Acceptable emulation simply isn’t an option.
It’s possible that Apple will find themselves in a position where they have no choice, and it’s also possible that they could pull it off. I’m not saying they can’t. I’m simply arguing against the cavalier attitude toward such a move that so many seem to have. Make no mistake–it could very well be a bet-the-company move.
I agree.
And it is based on the supposition that the PPC will never get speed increases.
And I firmly believe that in around (or less than ) two weeks, we will see that there is going to be a massive speed increase across the board that’ll make coursey et al’s articles moot.
If apple started making OSX to run on their x86 hardware, I doubt they would succeed anywhere in the market. Because not only would people hack the number system out of the software, they would also be forced to try to beat x86 computer makers like dell, gateway, hp/compaq, etc. Apple won’t have a chance. Now if they produce OSX as a boxed OS that users could purchase, install, and replace windows with. They have an actual chance at removing MS’s dominance on the desktop. Particularly if they get OEMs interested enough to replace Windows, and excite the BSD and Linux crowds enough. But if they are going to do anything like this, they better commit early, and get some prototypes out there with development tools, so that porting big name applications will be possible, and second so that big name companies will know that Apple really is serious about switching cpus and architectures. The Core of OSX, BSD is remarkably portable, and it makes perfect sense for them to move to a newer faster platform. But Apple has to get out of this exotic hardware business, because it is obvious that it really has kept the average joe from experiencing true innovation. They are basically shooting themselves in the foot by never even attempting to become the standard desktop on home pcs. They would rather make the cost so absurdly high that windows and x86 hardware in general looks like a deal. And it is sad that Apple wouldn’t take the opportunity to win back many of the customers they lost so long ago. The opportunity is now, the capability is most definitely a possibility, and the only thing left is to see how successful Apple really wants to become.
I think choosing the intel archicture for apple is an almost win-win situation, while keeping
the closed platform approach it has now, why?
1) To the end costumer whats the big diference between x86 ou powerpc??? the majority
os Mac users don’t know or care whether their computer is based on x86 or powerpc as long
as it works the same way it did!
2) Macs strong points are diferentiation, integration and design. None of this points is affected by the move to x86!
Macs will remain a diferent kind of computing experience for the user while using the same hardware has pcs.
Macs can achieve the same kind of integration is has currently by keeping the platform closed and wisely
choosing the hardware they want to support! The design can be just the same! No beige boxes for Mac x86.
3) Performance-wise x86 macs will for sure be better and faster machines than powerpc macs, at a very smaller cost
and at a very small risk that their supplier decides to stop producing new chips, or isn’t interested in developing them more!
4) The choice of hardware will increase dramatically and Apple can choose from a larger variety of suppliers what they
want to support while keeping their number of choices very small.
5) Macs can be PC convertible so the “risk” of buying a different platform is mitigated by the fact that it can be
converted to a regular pc ( if they want to!).
6) Sofware wise all those Multimedia apps that exist only on the pc and that use x86 optimized routines can be converted
much more easily to the Mac ( where they belong!??). The digital hub becomes more easily a reality!
7) Macs will probably become cheaper machines and so become much more mass appealing.
Like any other modern operating system, OS X is abstracting the hardware and nothing of the upper layers is communicating with the hardware directly. All the hardware abstraction is done in Darwin.
Do you know anything about XNU at all? I mean seriously, XNU is hardly a Mach kernel. It’s monolithic, has no Mach servers. The only Mach components left are the process manager and VMM. Maybe you should read up a bit before talking out of your ass.
CoreServices WindowServer is “communicating” directly with your display hardware.
The reason for apple to switch is largely processor speed. Yet the value of increments in processor speed continue to decline while the focus on usability, stability, and even low power consumption are increasing. Will anyone other than a few geeks really care if an apple runs at 4 GHz while intel chips run at 8 GHz. The ability to differentiate one’s platform based on speed is a thing of the past that most PC advocates just don’t understand.
now lets talk about the cost of an upgrade to x86 or any other non powerpc platform in the short term.
1) All applications would need to be ported, raising a high potential for limited applications and we know what that does
2)Apple’s developers would be pissed beyond belief. Why because an increase in market share can be achieved by simply lowering apple’s pricing scheme without any need for ports.
3) Apple’s user base, you know those people that spend $2000 on their snazzy desktops, would be furious.
4) Apple would also confuse users
5) apple would lose its marketing edge.
6) Apple would find it difficult to price products at a premium when they are selling the same thing as dell.
Are there easier ways to attain a better performance or higher processor speeds? Yes yes yes. It would be FAR easier to get someone else to fab the G4 or G5 than to switch to x86. for example
1) IBM. IBM continues to show a lot of interest in expanding their role in the processor/semiconductor world. They are also involved in teh cell project with toshiba/sony, and they remain a supplier of G3’s
2) AMD would also benefit from higher volumes and the ability to better utilize each generation of chip making material. Cap ex for chip making is enormous. The more customers, the merrier.
In addition,
the assumption in all of these switch to x86 arguments is that Intel’s way is the correct way. not necessarily so. By being the first to switch to a new generation of chip making gear, intel effecitively subsidizes the costs. Intel is leading the speed curve because they know that it gets harder and harder for a lower volume producer to justify keeping up with them. but in the process, they are making faster processors less relevant to the masses. There are also better ways to get peformance and as the register put it most of the adult world is proceeding with that.
You can turn Intel’s strategy on them in two ways.
1) proceed to get higher market share (makes intels speed game tougher to justify). I believe apple can do this just by lowering prices. They will need a high end though to justify that.
2) get performance in other ways, such as coprocessors or greater vectorization and reap the benefits of intel’s subsidization. EVen microsoft sought to achieve performance on the xbox by using a dedicated video processor.
3) acquire sales based on other features such as usability or diversity. processor speed is not the feature that is was in regards to sales.
A short term switch to x86 would be a disaster for apple. Still, four years is a long way off and anything could happen then. x86 is not the only choice though. Apple has, as mr jobs put it, “options.”
One way would be to to just port OSX to std x86 AND just gauge for it to make up for not being able to go head to head with Dell etc. Right now it goes for $90-$110 IIRC, lets call it a subsidised OS by the overpriced ppc HW.
So how much would you pay for OSX x86 edition giving Apple fair due for loss of HW sales. I would probably go about $300, the same ballpark as say VMW or the full price for XP or other high end SW people have to buy.
Any less & I know I amy getting a steal, any more I would have to be sure its the right thing.
This way they can gradually slide out of the HW business but make a handsome profit on the OS. I don’t even mind if they use a dongle as long as it works with any PC I use, say a USB rom. I would want to be able to install on a few machines, & just move the dongle. Even better if removing the dongle just suspended the OS so as to keep state while I use another box.
But I want control of the pc HW, & I want to be able to multi OS & any Apple x86 will probably screw that option. Just better give me uber protected SW for fair price.
Well it won’t happen like that, but
“I would probably go about $300, the same ballpark as say VMW or the full price for XP or other high end SW people have to buy.”
You ever think that MS might just be ripping you off?
Maybe you should read up a bit before talking out of your ass.
Gee. You better stop using four letter words if you want me to talk to you. Thank you.
OK. I should read it up a bit. Does Darwin source code count? Compiling self-hacked IOKit components?
CoreServices WindowServer is “communicating” directly with your display hardware.
Really? How come it is identifiying graphics cards by the (editable) properties in the device tree the IOKit is showing it instead by their PCI IDs?
And even if it was: Anything that can be hacked will behacked.
Fade out Macs like the Apple ][s were faded out.
Come out with ‘Platform x’ that uses Sparc CPUs (or other non-x86); continue to market Macintoshs and Platform xs BOTH, until the Macs just can’t keep up (sales wise, not speed wise).
How about Apple create a ‘Live Filesystem’ bootable CD that is compatable with x86? Runs 800×600 VESA compat. and supports the usual NIC suspects.
That way x86 users can try and if they like then buy Apple hardware and make the change.
My reason for this, I am a Linux/BSD/Windows user. I would love to have a Mac running OS X. Slight problem with budgetry constraints 😉 and also the unknown (that being OS X) I would love to take it for a test drive just to confirm that spending the money would be a good thing. So far, my only option to feed my need to know is to rent an iMac for a week with OS X loaded …
If Mac was really going to do this sort of thing they would have done it when Windows was widely thought of as unsatisfactory, not at a time when everyone is waxing poetic about how stable the recent iterations of Windows are. This is just a scam by Mac to scare Motorola into concessions we don’t even know about.
The arguments for switching to x86 are indeed quite compelling if Apple was starting from scratch. But alas this is not the case. They are still trying to switch from Mac OS 9 to Mac OS X and you think Apple can say “ok forget everything we said about Velocity Engine, Photoshop optimizations, beautiful risc vs their kludgy cisc, etc. Developers just port your apps to x86!” Apple is quite stuck and really depend on a miracle from Motorola to save them now. They can concentrate on consumers with the iMac where speed doesn’t matter so much but if they really care about being some mediaOS by buying out emagic, etc then speed definitely matters!
Next time you run a feature on OS/X86, how about put a link to the comment pages on the prior feature.. save us all repeating ourselves…
( I’m sure I’ve seen some of these comments before.. nuffink new here.. )
Definitely something I would buy. (No irony, I mean it)
A hardware switch to Intel from PowerPC would be suicide for Apple – because when all’s said and done, Apple are a hardware company – OSX vs WIndows on commidity hardware (and it would happen eventually no matter how Mac specific apple try to keep the hardware spec initially) would be the same as pitching BeOS vs Windows – with the same result
For those who have Macs, remember that Apple has done the processor change dance before. It was not as drastic, mind you, but it happened. For instance, the Switch from the ‘040 to the PPC architecture was significant in Apple’s past. A compatibility layer had to be enabled in the MacOS for the 040 apps to run on PPCs. PPC native code ran much faster. That shift forced everyone who bought a new Mac with a PPC to either run the old apps in emulation mode or buy the new PPC version.
Admittedly, it would be quite different to switch to x86 from PPC. Emulation mode would be possible, just as there is software available to run Windows on the Mac. A VMWare-like software layer, running either OS9 or OSX PPC could be possible, but would take time to develop.
The porting of software by the various companies would not be that bad, I think. Afterall, Adobe and Quark products are already available for x86 Windows. So, optimizing for SSE, etc. has been done. The next step is to combine the OSX GUI API interface with what they know about the x86 processors.
Yes, Apple users would have to buy new versions of their software, but it would be something they have done before. Windows users have the same issue every time a new version of software comes out. The bummer is that just like with the PPC conversion, it would take time for the app companies to make the switch.
The trick might be for Apple to talk the app companies into allowing customers to exchange the same-version software. In other words, give us your PPC version, and we’ll give you the x86 version for little or no cost.
As for Apple preventing OSX on generic x86 hardware, I bet they would use the proprietary ROM chips again like on the Macs. If they modify the BIOS or require onboard ROMS of some sort, then other x86 PCs couldn’t use the OS (without some hacking).
As for the time it would take to make OSX run on x86… Darwin already does, and if I understand correctly, that is the non-GUI base of OSX. So, if they tweaked and recompiled everything from the GUI layer onto x86, they would be ready (I say this as if it were simple, but it might not be).
Hardware would be easy. Apple’s MO is to select a very small base of hardware – sound, video, etc. So, drivers would not take too long. I bet Darwin already has most if not all of the drivers necessary (aside from 3D video). Since Apple has already used ATI and NVidia as OEM partners, they would probably continue to use them for the x86.
A.N.
If they do use x86 it will be from AMD because they have an easy migration path to 64bit. The last thing Apple wants to do is get stuck with 32 bit, a la, P4, or stuck into 64 bit, a la, Itanium.
Of course there is alwas Yamhill.
What if, instead of porting directly to x86, AMD created an athlon or hammer or opteron chip without the x86 overhead. Underneath, all pentiums and athlons are essentially RISC processors, with a bunch of translation to get the x86 code to work. Drop the x86 part since you’re making an architecture switch anyway, and suddenly you get a ton of previously unavailiable power. You’re running on almost x86 hardware, and you benefit every time x86 gets faster, but it can never catch you, since it will always have the overhead.
That shift forced everyone who bought a new Mac with a PPC to either run the old apps in emulation mode or buy the new PPC version
The old apps ran great in emulation mode. That’s the point; the users’ existing investment was not invalidated. Without that kind of mechanism, a transition is excruciatingly hard. The problem, of course, is that running 040 programs on a PPC was waayyy more acceptable than running PPC programs on an x86 would be. The x86 is faster, to be sure, but not by the margin of 040–>PPC.
Yes, Apple users would have to buy new versions of their software, but it would be something they have done before. Windows users have the same issue every time a new version of software comes out.
Normally, you can choose to upgrade to a new version of software if the new features are compelling enough for you. A switch from PPC to x86 would mean that you would either have to upgrade or accept the fact that all of your applications no longer work. Upgrading for new features and upgrading to continue working are completely different concepts, and I’m a bit surprised that you even mentioned the former (it’s completely irrelevant to the issue at hand). And, for what it’s worth, with the exception of some antivirus tools and a few other specialty utilities, Windows has rather amazing backwards compatability. Customers don’t tend to like it when suddenly all that they have becomes useless.
The trick might be for Apple to talk the app companies into allowing customers to exchange the same-version software. In other words, give us your PPC version, and we’ll give you the x86 version for little or no cost.
That would certainly make most customers happy enough, but it’s not going to be popular with the developers. Basically, it would be, “We appreciate your supporting us here at Apple, and we appreciate your Mac software. Now, because of the decisions we’ve made, we’re going to ask you to port your application to the second new codebase in three years. Oh, and to be nice to the users, don’t charge them for it. You should shoulder the cost of development entirely. Thanks again for being a Mac developer!”
Development is anything but free, even for “relatively simple” ports. So either you piss off your users or you piss off your developers (or worse, both). Don’t make the mistake of underestimating the challenge (and danger) of such a move.
“Drop the x86 part since you’re making an architecture switch anyway, and suddenly you get a ton of previously unavailiable power.”
the x86 decode part of the chip is very small .. it used to be relativly larger. Removing just this section would only create a little more room for transistors and certainly not “a ton of previously unavaible power”. The ATH MP has 80 million transistors and the p4 110 million.. this shows that its not even just the num of tansistors that matters but the design.
The Athlon is really a redesigned DEC alpha (because the same engineers made it) which is why it performs so close to it.
The way out for Apple is to have a dual cpu computer 1 g4 and 1 P4.
I find the hardware cost issue with Apple to be really overstated. I recently bought the new iMac for my wife and for me to try out OSX longer term. For $2000 I got:
17″ high end lcd
dvd-R/CDRW
these two components are worth well over $1000.
For the other $1000 I got:
G4 800mhz (call it a pentium III 1ghz)
512 megs ram
80 gig hd
32 meg Nvidia video card
firewire and usb support
speakers
network, modem
keyboard, mouse, case….
Top notch OS (call it Win XP Professional)
That doesn’t really seem that expensive to me. Try and put a system like this together for $1000. When I do it it comes in over $1300 every time.
The powermac line will stop being overpriced next month; apple doesn’t do monthly price changes like PC manufacturers.
Is this OSNews.com or is this IntelMacs.com? Is it now OSNews policy to have one good PowerPC-is-dying-and-Apple-needs-to-move-to-Intel discussion per week? Has there been a single sentence in this thread that hasn’t been puked up in past threads? What else is left to discuss? Is anybody else getting tired and annoyed by this?
It has become a broken record I’m afraid!
it is a manifestation of the PC’s crowds deeply rooted Mac envoy syndrome. They want a mac but won’t pay for one and it obviously eats away at them.
sorry, not likely to happen… amd and intel’s mobile procs run too hot… and are too power hungry…. apples powerbooks last a long time on batteries…. and they don’t burn your legs when you really use it as a “laptop”… not unless some serious power issues are taken care of will this ever happen.
The problem, of course, is that running 040 programs on a PPC was waayyy more acceptable than running PPC programs on an x86 would be.
It is true that the old apps ran fine. But, most of the Mac users I know don’t really care what the processor is. They use the OS because its use requires little technical knowledge and tweaking to do work.
The x86 is faster, to be sure, but not by the margin of 040–>PPC.
PPC emulation would be a painful way of running old code, but it’s the only way I could see the PPC apps having any extended life. Besides, x86 processors get faster at a higher rate than PPCs. It may not be long before x86 is much faster than PPC.
Upgrading for new features and upgrading to continue working are completely different concepts, and I’m a bit surprised that you even mentioned the former (it’s completely irrelevant to the issue at hand). … Customers don’t tend to like it when suddenly all that they have becomes useless.
I don’t see the two upgrade paths as irrelevant. It’s true that the two upgrade types are not the same, but my thought was you would keep using the PPC version with your old Mac unless you decided to go to x86. At that time you could buy a new version of the app as well. More expensive, yes (like Apple products in general). But it’s a choice. The ideal path would be for the app companies to produce versions for both platforms for an interim migrational period.
That would certainly make most customers happy enough, but it’s not going to be popular with the developers.
It’s the develpers’ choice. They can continue to produce apps for the “legacy” PPC platform or they can migrate to the “new” x86. This is similar to the OS9 to OSX change. You can run apps in the OS9 window now, but they may not work as well as in native OS9.
Besides, if Apple makes the choice to go to x86, its ultra-loyal users will be forced to follow as they upgrade. Apple moved from ADB to USB. Apple users scrapped the ADB peripherals when they moved to new hardware. They moved from SCSI to IDE. Apple users had to buy a SCSI card for new Macs or not use SCSI at all. Apple dumped the floppy. Apple users were unhappy/pleased. Apple is moving to OSX. The trend is: Where Apple goes, the loyal users and Apple app makers follow.
Any Nonmouse,
Though bkakes and I don’t see eye to eye on certain issues, his analysis on this subject is right on in my opinion!
Is this OSNews.com or is this IntelMacs.com? Is it now OSNews policy to have one good PowerPC-is- dying-and-Apple-needs-to-move-to-Intel discussion per week?
Amen. I’m tired of people being constantly reminded that their G3/G4 Mac systems are horribly inferior to their x86 counterparts. My response is:
So what? It’s nice hardware with a nice operating system. It doesn’t generate a great deal of heat and it’s relatively quiet.
is that consumers can make DIRECT and UMAMBIGIOUS comparisons with other vendors on price/value axis.
if apple on x86 charged significantly more than the competitors, and since direct comparisons are now possible,
it may end up loosing market share.
so whether a 800mhz G4 imac2 is comparable to a 2000ghz P4 is a matter of speculation. but a 200ghz apple p4 is directly comparable to a 2000ghz dell p4.
apple needs higher margins because unlike dell, apple needs to channel profits to r&d.
dell does not develop new technologies. it is not a technology company in that it has a portfolio of patents that generate revenue. it can undercut apple.
stew: First of all, see how Apple’s current marketing strategy draw focus away from CPU speed to user experience. They don’t give you any tech specs, numbers or anything; It’s just “you can do photos, movies, music, email and web on it”, and the current G4s are sufficient.
No, it’s not sufficient. Mac OS X is very complex, and its power can’t really be realize on PowerPC processors. Just thing of it, things would be faster by 2-4 times. Photo editing on the G4 is fine, by I notice removing red eyes from photos in iPhoto isn’t as fast as on Ulead’s on the PC.
stew: Second: Apple will not be able to tie OS X to their computers with Darwin being open source. They already failed on the PPC
Surely, there isn’t a secure way of making Mac OS X Apple-only, but really, how much of Apple’s target market knows how to modify Darwin – or even know what’s Darwin?
ckristian: Maybe if they move to AMD Hammer .. they still need to find a way to say “this is a mac not a PeeCee”
To the consumer’s eyes, the difference between a normal PC and the Mac is the OS. On the technical stand point of view, the difference between the two is the chipset and the processor.
Besides, if Apple could rename the G4 as G4, as oppose to 75xx, why can’t Apple rename Hammers or Pentium 5s?
So explain to me, why is it suicidal? To the consumer, the difference between 1GHz and 2.53GHz is very big. Now, if tommorrow Apple releases a 2.53GHz Pentium 4 based system, it could compete much better with Dell’s P4 offering.
BTW .. all the optimization work done for years by developers on PowerPC ?
If Apple had the same mentality when it moved from 68k to PPC… I wonder where they would be.
Maybe another emulation framework? running PPC code on Hammer? how many years to get it working properly?
Apple might in fact be working on it right now. Besides, how long did Apple took to write 68k emulation when it moved to PPC machines?
If they can find a good technocal solution ok .. but .. maybe if I want an Hammer based computer I will go for a PC clone … why go for apple?
The very same people who go for Apple right now: people who want Mac OS. Notice Apple’s main selling point lately had never been the processor, I don’t this changing.
CattBeMac: It’s only PC/Mac users with common sense know that Apple going Intel would be the death of Apple!
Amazingly, CattBeMac, you never explained why.
bn: Intel will have that much more of the overall CPU market. It’s sad.
The most they would get is an additional 2.4% of the world market, big deal. It isn’t like Intel noticed the PPC market nowadays. As long AMD survives, there would be sufficient competition.
Richard Fillion: They have an appeal to them. I think a lot of that appeal would be gone if they’d just to x86, it wouldnt be exotic anymore
It seems that the appeal comes from the looks and how it works rather than the processor it uses. I don’t see why Apple’s hardware should look like normal beige boxes once they use x86 processors… there isn’t a law to make them do that.
john/*: It has 2 1800 or 1600MHz G4’s. […]
Such a thing is not in Motorola’s roadmap. Dream on. (And even if it is, at the rate its going, by time they actually come out with them, we would have 6GHz P5s..)
ActiBurro: Windows will run over this x86 box and a lot of user could decide go to Windows
If they want Windows so badly, why would they buy a Mac? Besides, Apple still sold a machine, they made money.
Bascule: they could build a system without an ISA bus entirely
Do modern machines come with ISA in the first place?
A more plausable scenario is going with all IBM PPC chips
After months of discrediting the G3, they are going to move to it because of a 300MHz difference?
Bascule: The difference is the iBook doesn’t have a fan, doesn’t make heat, and has much longer battery life.
My laptop has a battery life of 5.5-6.0 hours, depending on what I do. Though, when I use Linux, it goes down to 4 hours. It also takes an hour to fully recharge the laptop.
Besides, as for the heat, you notice your Palamino emits less heats as the Thunderbird. Thoroughbred and Palamino has the same pattern, and the same should happen when Barton and Clawhammer comes out.
bkakes: All software would have to be, at the very least, recompiled.
This didn’t kill off Apple when the 68k-PPC migration happended. There was something called emulation.
bkakes: They would risk decimating their brand
How? They still sell Macs, only with Intel/AMD processors.
bkakes: The rather painless move to PPC from 040 was possible because they not only provided remarkable emulation, but also because the PPC was so much faster than the 040.
In 18 months, x86 processors should be 2 times faster than what they have now. It’s called the Moore’s law.
john/*: And it is based on the supposition that the PPC will never get speed increases.
Rumours say that Apple is release 1.2GHz PowerMacs, while Intel is releasing 3GHz P4s. If the patern continues, by time Motorola reaches 2GHz, Intel would have past the 6GHz line.
Apple has a major reason not to switch to an x86 chip. Their hardware sales make up 80% of their profit.. you take their platform to x86, all evidence points that it won’t make the headway that anyone would hope except for those currently running mac looking to run it on cheaper PC hardware.. which will take away from their hardware sales, which is what keeps them floating. Not going to happen..
Read this one. It makes sense.
If MaC Os X were migrated to x86, surely it would execute only in a special x86-box, but in a few months a hacker would make it executable in every x86 machine, and Apple would end shelling an OS an a few more.
“The way out for Apple is to have a dual cpu computer 1 g4 and 1 P4.”
Obviously not. Apple had done this before, way back when 486 was hip…
with darwin being “opensource” and ported to x86 already it would be little trouble to port the kernel back without the need for the custom chipset that apple would almost certainly use if they were to start on x86. whatever happened to the osxonintel site? optimising Photoshop for OSX on intel would be far easier than optimising for G4 is as the compilers already support the latest x86 enhancements etc.
Someone above mentioned taking an AMD an removing the x86->RISC tanslation unit. It’s not quite that simple as you need to add a PPC->AMD-RISC translation unit and probably do a load of other internal changes but it might well be possible.
So you would have a nice fast x86 except it wouldn’t be an x86 – it would be a PPC. This would take time and money but at least wouldn’t require emulation and all the other complexities compared of switching to x86.
On the other hand they could do some work on the compiler (they are) do some work on optimising the OS (they are) add a decent speed bus (they are) put the G4 on 0.13um boosing it’s clock speed (they are) they could improve performance that way.
Once Motorola to get their finger out and produce a G5 I think these speed issues will be moot.
On the other hand they could always do this if they really wanted speed:
http://www.theinquirer.net/?article=4842
I would love to see this happening. I would love to run a MacOS X on my p4 2.0 machine. Imagine how it would be going to the nearest computer store, and MacOX X, MacOS X on the shelves everywhere, just like you have Windows XP. I want to run MacOS but it’s too expensive for me, I can’t afford it, but I will buy their MacOS X from the store if they make it avaiable, This is like, a dream come true, I really like MacOS X. It will definitely mean good bye Windows XP for me.
Isn’t it far more likely that Apple would develop their own mainboard chipset that uses x86 or x86-64 processors, sdram, ide, agp, and pci, but is NOT compatible with PCs? That way they could keep their hardware business and reap the benefits of x86* processors (low price for performance and availability).
Chipset, processor, sdram, ide, agp, and pci. That’s a PC right there, doesn’t matter who makes it.
Motorola did make the G5 and it’s supposedly up and running in their labs. The problem is that Motorola is half-assing it. Apple really is a go-getter company always pushing forward, and it’s partners like Mot that slow Apple down. For Apple to advance at the rate they want, they have to find a partner that will move as fast as Apple is. x86 CPU development moves the fastest and is the most competitive. Even lesser companies like VIA and Transmeta are producing x86-like CPUs that run Windows.
The lack of speed and performance in PPC chips isn’t because of poor design. It’s the lack of interest that’s dooming the platform. If Motorola would wake up and realize that OSX is a viable competitor to Windows, they might take it more seriously. Until that happens, Apple has to keep moving forward and decide whether to wait for Motorola to come around, or start looking for a new partner.
from news.com
“Intel’s Banias: Not built for speed”
http://news.com.com/2100-1001-948791.html?tag=fd_top
Shows even intel will start moving away from pure Pc speed, at least for notebooks. It is a worthy read.
I don’t see Apple moving to X86 anytime soon. It will be to hard and to much of a hassle. First they would have to get all of their exsisting customers on OS X then they will have to convince all of the Mac developers to write “new” versions of their software. It doesn’t look feasible, but what does is the new Power4 that is designed for the desktop/ entry level servers.
http://www.mdronline.com/mpf/conf.html#day1_2
Here’s the blurb….
Session One: PC Processors
Kevin Krewell, Senior Editor, Microprocessor Report; General Manager, MDR
Breaking Through Compute Intensive Barriers – IBM’s New 64-bit PowerPC Microprocessor
Peter Sandon, Senior Processor Architect, Power PC Organization,
IBM Microelectronics IBM is disclosing the technical details of a new 64-bit PowerPC microprocessor designed for desktops and entry-level servers. Based on the award winning Power4 design, this processor is an 8-way superscalar design that fully supports Symmetric MultiProcessing. The processor is further enhanced by a vector processing unit implementing over 160 specialized vector instructions and implements a system interface capable of up to 6.4GB/s.
An interesting note. IBM has gotten out or is getting out of the whole desktop arena. They are still in the whole server thing, but this is also aimed at the desktop market. Unless I’m mistaken only one computer company would uses a a PowerPc based processor. Apple. This looks like the replacement of the G4
The underlying assumption behind this x86 Mac rumor is that Macs would be better if they were more like PCs. I don’t know whether that’s technically true, but in terms of creating a firm brand identity for Apple it’s demonstrably false. “We’re just like PCs now!” is not a winning ad slogan. Future Mac models may very well be based on a chip other than PPC, but the new hardware will have to be something that Apple’s marketing department can plausibly argue is superior to the competition, not identical.
>>CattBeMac: It’s only PC/Mac users with common sense know that Apple going Intel would be the death of Apple!
Amazingly, CattBeMac, you never explained why.<<
That is because I would just be repeating myself again (no use in beating a dead horse)! Besides, bkakes summed it up quite well!
of the following statement appearing in a thread about Apple going x86…
>>no use in beating a dead horse
-bytes256
Do modern machines come with ISA in the first place?
Yes, the ISA bus is where things like, oh, your BIOS, the 8042 controller, and depending on your south bridge serial/parallel ports are connected. Basically, any legacy crap left over from the XT days.
A more plausable scenario is going with all IBM PPC chips
You mean POWER?
Besides, as for the heat, you notice your Palamino emits less heats as the Thunderbird. Thoroughbred and Palamino has the same pattern, and the same should happen when Barton and Clawhammer comes out.
Average operating temperature for a lower end Palomino is ~39C. That’s a hell of a lot higher than room temperature.
I have two Athlon XP 1700+’s (1.53GHz) and they both operate at 46C. The heat problem isn’t going to get fixed any time soon.
No matter what you say about cooling, you always run into this problem:
1. Quiet
2. Effective cooling
3. Inexpensive (i.e. not water cooling)
Pick any two.
The G3/G4 let you have all three.
Drop the x86 part since you’re making an architecture switch anyway, and suddenly you get a ton of previously unavailiable power.
Umm, no. You’re talking about getting rid of the decoder. What you’d be left with then is a processor that takes micro ops specifically designed with the intention that they were being decoded from x86 opcodes. Furthermore, you’d have to know which execution unit they were intended to go to. You just can’t get rid of the decoder, it’s an integral part of the processor.
And even so, getting rid of it wouldn’t exactly offer you any advantages over having it there in the first place (unless it’s a first generation Pentium 4)
The Athlon is really a redesigned DEC alpha (because the same engineers made it) which is why it performs so close to it.
Umm, no. It uses the same bus architecture, and that’s where the similarities end. This is sort of true if you consider that Digital invented the basic concepts behind modern superscalar processors.
I forget the name and don’t care to look it up now, but AMD bought some company who came up with the original design for a superscalar x86 processor.
The way out for Apple is to have a dual cpu computer 1 g4 and 1 P4.
The hell? No, no, no…
I don’t see how going x86 would hurt Apple’s hardware sales since they are the only company that build and distribute Mac systems. And the idea that a hacker would enable this hypothetical x86 version of OSX to run on any non-Apple x86 hardware, is not as significant as most people think. What’s stopping a hacker from making the current OSX run on non-Apple PowerPC platforms, sort of like reviving the clones?
The reason why no one has done it yet – and this is just my guess – is because no one cares. I would doubt a “hacker” persona would appreciate the Aqua interface therefore have no real incentive to hack OSX to run on open platforms.
Really now, the die-hard Mac fans buy Macs because the case looks nice and they can brag about it to everyone they meet. Those kinds of fans don’t care if the processor is dog slow, they’re mesmerized by the beauty of the friggin nutshell. So Apple will still get sales in hardware from people who like the way the boxes look.
I associate the pretty looking case to the stereotype of “dumb blondes” – I apologize to all you blondes out there. It looks nice on the outside, but inside it loses all that glimmer.
http://www.theinquirer.net/?article=4862
bkakes: All software would have to be, at the very least, recompiled.
This didn’t kill off Apple when the 68k-PPC migration happended. There was something called emulation.
I would love to see where one can find acceptable PPC emulation on the x86. (There are all sorts of articles written about why it’s so difficult.) The 040–>PPC change was huge, and that’s why such successful emulation was possible. And there’s the other issue that no one seems to note: the 040–>PPC migration started back in 1994, when Macintoshes still enjoyed a huge advantage over Windows in many areas. Whether or not you think the Mac is better today, the margin certainly isn’t nearly as large; Apple enjoyed more leniency back then.
bkakes: They would risk decimating their brand
How? They still sell Macs, only with Intel/AMD processors.
Easy, exactly as I explained: They would no longer sell simply “Mac”s. They would have two different–and incompatible–lines: “Mac – PPC” and “Mac – x86”. Assign whatever friendly names you want to these product lines, but the problem remains: you’ve splintered one line, one brand, into two. That’s incredibly damaging, especially when much of your market is people who don’t want to deal with computer hassles.
Much of the appeal of Macs is that it’s a happy little world where in theory everything works together nicely. But PPC software simply won’t run on x86, and the same is true for much hardware, such as video cards and other endian-dependent devices. Apple is having a difficult enough time getting people to migrate to OS X–and believe me, there’s a reason that OS X looks very little like OS 9–can you imagine if people had to look at the box for “requires a Mac, OS X, x86 version”?
A lot of people here seem to think, “They can drop in a different processor and it won’t matter to the end user; they’re still buying a Mac.” But that’s completely wrong precisely because the software and some hardware is incompatible. Knowing which type of processor your Mac had would be required knowledge if you wanted to buy anything, and that’s where the brand would suffer so much. Splintering your brand has major consequences. Can you imagine if there were two type of PS2s? Two types of DVD movies? You get the picture.
bkakes: The rather painless move to PPC from 040 was possible because they not only provided remarkable emulation, but also because the PPC was so much faster than the 040.
In 18 months, x86 processors should be 2 times faster than what they have now. It’s called the Moore’s law.
Back in the 040–>PPC move, the PPC emulation could run often run 040 code faster than the fastest 040 machine. So let’s assume that in 18 months there’s a dual 1.6 gHz G4. This means that in order to match that speed in 18 months, there must be an emulator that can, right now, run PPC software on an x86 as fast as a single-processor G4 1.4 gHz or so. Can you please provide me with the URL?
Hmm. I remember saying this in the last thread–that the most likely candidate for a new Apple CPU would be a new PowerPC chip based on the Power4 architecture.
For all the reasons I and other people have talked about, from marketing perspective to maintaining binary compatibility (anyone who thinks that Apple is going to make a that requires ISVs to recompile all their applications, particularly when the push to get people to move to Cocoa or at least Carbon is still underway, needs to cut back on their bourbon intake a little), this just strikes me as a much more likely alternative than a move to x86.
as quoted;
“Breaking Through Compute Intensive Barriers – IBM’s New 64-bit PowerPC Microprocessor
IBM Microelectronics IBM is disclosing the technical details of a new 64-bit PowerPC microprocessor designed for desktops and entry-level servers. Based on the award winning Power4 design, this processor is an 8-way superscalar design that fully supports Symmetric MultiProcessing. The processor is further enhanced by a vector processing unit implementing over 160 specialized vector instructions and implements a system interface capable of up to 6.4GB/s.”
Looks like Apple will more than likely jump on this and fuse together a nVidia/IBM/Apple relationship if Apple does in fact utilize nVidia’s latest techno with their chipset and mobo configs!
The Intel/AMD (and anything else x86) will never make its way inside any Mac, and I mean NEVER!!!
>>Really now, the die-hard Mac fans buy Macs because the case looks nice and they can brag about it to everyone they meet. Those kinds of fans don’t care if the processor is dog slow, they’re mesmerized by the beauty of the friggin nutshell. So Apple will still get sales in hardware from people who like the way the boxes look.<<
Now you know better than that… I see more PC folks bragging about what they just slapped together on their homebuilt machines and their triumphs over technology than some Mac user bragging what his machines looks like. Grant it PCs are usually very ugly and don’t have much appeal, but I remember when I saw the first iMac I laughed myself silly (I was a dedicated PC user at the time), but then when it was time to upgrade my old PC, I had all I could take from Windows annoyances (Windows NT pushed me to the edge when I got my job in Europe), so I did some research and went the Mac route. I didn’t really think about the pretty box, but I was more impressed with the track record of Macs in general where reliability was concerned. And when I have been so spoiled by Sun Microsystem’s hardware/software solutions and how reliable they were from working around them for so many years, that was the kind of experience I wanted (of course I couldn’t afford a Sun D’OH!)!
And from a historical standpoint PCs were always slower than Macs for years. Just in the last year or so the Mac has been lagging behind the PC, and we all know why that is… Motorola!
Well I have been screaming(not here) for years that IBM road map was promising and Apple should use there PPCs chips. I do see Apple using the new announced IBM PPC chip and FORGET Moto.
If you read these posts, all involved pretty mch agree Moto has screwed Apple at every turn.