Apple would have to give 3rd party developers at least 6 months advance notice of any architecture change. What good is an OS with no apps?
Not to mention all us developers would be POed! Such a move would drive us into WinXP faster than a meteorite landing on cupertino.
Stop passing the hot turd around like it's a box of delicious chocolates.
Um... Er... No.
The point - the POINT of going AMD is not a platform change, but a platform add.
Apple can, though Cocoa (sorry carbon kids), add AMD64 to their PPC line and push out not only to some new markets, but also buy a hedge against a sole CPU supplier.
Provided that Apple can bring their software forward to generate profits, Apple can trim hardware margins to stay competitive and be a first-to-market unix-but-not-unix 64 bit OS vendor.
AMD would be a new market, not a replacement. Mac OS X may become that which NT could never be.
The use of AMD chips by Apple would amount to suicide for the company. Unless, of course, it were part of some other strategy to weaken Intel and strengthen IBM.
Apple could do that, and also get into places that don't even look at hardware that isn't x86. Sadly, such places exist. And if it introduces some more confusion into the PC world, so much the better.
I don't think anyone is talking about Apple leaving the PowerPC at this point. Yes, there was talk of that sort during the 500MHz debacle, but with that behind us and a good future mapped out for the next few years, I can't believe Apple has any intention of leaving their baby behind.
The question is, what would Apple release that would be based on an Opteron? I don't really have an answer to that, beyond "a server?" It sure isn't going to be powering iMacs...
AMD would be a new market, not a replacement. Mac OS X may become that which NT could never be.
And exactly who is that market??? Say that Apple releases a new PM with the Opteron, and even if they release all their iApps, Server software aso, there is almost no applications ready for this market yet. So they essentially would have a very fast and competitive computer with no market. I don't see very many professionals buying macs that cant run any of Adobe's, Macromedia's, Microsoft's or Quark's software for another maybe 12 - 24 months. Unless maybe Apple makes some sort of emulation package available.
And even if the Opteron is showing even better performance than a 1.8 ghz 970, the 970 is still a great processor and has already proven (if IBM's press release is something to rely on) capable of even greater speeds. And we are only 12-24 (hopefully) months away from an even greater processor - the 980. Before that we'll also have the choice of the 970+ so the future is not completely dark
There is also another question. What about portables? Will the Opteron be a good choice for laptops heat and battery-vise? I think not.
I guess I don't understand the logic of Apple making use of the Opteron or Athlon64.
Neither do I. I don't think Apple is going to switch from PPC anytime soon.
Quote:
If Apple wanted to make a move over to another architecture, the Itanium would be a better chip.
[Rimmer]
Wrong, wrong. Absolutely brimming over with wrongability. [/Rimmer]
Itanic is very expensive, and despite what you have said elsewhere, Intel can't price it to compete with G4, 970, x86-64, IA-32 or even Alpha. This would be Apple's death blow for sure. X86-64 would be a much better choice (which I don't think will happen - see above).
Quote:
Apple could port OS X to the Itanium and ride Intel's ability to scale the chip. In my estimation, this would be a much safer route as Intel has shown that it can consistently deliver the results.
The chip will never get cheap compared to PPC without major performance sacrifices which it can't afford since it doesn't have much of a performance lead.
Quote:
We should be content with IBM's new found committment to the PowerPC. Apple should do everything it can to encourage IBM's ongoing dedication to the chip.
Yes, yes and yes again.
Quote:
In any case, OS X on the Itanium is within the realm of possibility. OS X on the Opteron/Athlon makes no sense.
[Rimmer] (see above) [/R]
What make far more sense for Apple, if it wan't to get into the high end (which is the only place Itanium is tageted in the current roadmap) is to start using Power chips and develop 4x, 8x and up servers, etc. Or port OS X to IBM's hardware and sell them together in partnership with IBM.
IBM is going to stick with PPC for a long time (as you imply). They will be commited to it in their blades and workstations (as a 604 and Power3 replacement) and will make new, faster ones. Power4 was the first step in IBMs refocusing on the fact that it had all the technology to develop the best processors but was futsing around with a team than wasn't given enough attention from the top. Now it has that attention. IBM is talking about the death of all 64-bitters except Itanic and Power. They even want to squash Itanic. They will stay in the game and give Apple what they need.
Who knows? Apple's market share might rise enough for Mot to get really committed to desktop and workstation design and we could see Intel and AMD in the PC world with IBM and Mot in the PPC world, all strong and healthily competing and giving us good products.
And exactly who is that market??? Say that Apple releases a new PM with the Opteron, and even if they release all their iApps, Server software aso, there is almost no applications ready for this market yet. So they essentially would have a very fast and competitive computer with no market. I don't see very many professionals buying macs that cant run any of Adobe's, Macromedia's, Microsoft's or Quark's software for another maybe 12 - 24 months. Unless maybe Apple makes some sort of emulation package available.
And even if the Opteron is showing even better performance than a 1.8 ghz 970, the 970 is still a great processor and has already proven (if IBM's press release is something to rely on) capable of even greater speeds. And we are only 12-24 (hopefully) months away from an even greater processor - the 980. Before that we'll also have the choice of the 970+ so the future is not completely dark
Well, Opteron is mainly targetted at servers. The market is those who would just automatically grab an x86 solution so they have unix/linux/Win compatibility. Apple could potentially build AMD powered Xserves, clusters, grid setups. Companies can confidently buy them with the knowledge that if OS X doesn't work out, they can fall back to traditional solutions. Apple doesn't need to deliver iApps here for a while. Only developers targetting servers need to jump immediately, and many of them deal with POSIX or Cocoa APIs, so that's probably a snap for them.
My guess is that Apple could bring products to bear on a new platform faster than anyone else including MS since Cocoa is designed to target multiple platforms. Anyone currently writing cocoa apps are likely not much more than a recompile away.
Since AMD/OS X could be 64bit only, it doesn't uncork the genie implied by x86 support, Apple could maintain their hardware control and reach into enterprise that want their MS backup. It's the fact that Opt and 970 are close in performance that makes this attractive for Apple since customers don't have to chase benchmarks to decide. It gives Apple a backup in case one or the other craps out. It gives developers a new family of hardware to go after, which is *good* since Apple's dev tools make it so freaking easy to support both, and it gives Apple the ability to lead on all leading 64bit platforms.
If Apple does want to go after a broader market and they want to go into MSs turf (compete head-to-head with Windows on the same hardware), they need some kind of pristine environment like this to jump into that doesn't suffer from legacy driver problems, and all of that. Apple could bring OpenFirmware to Opteron, do away with BIOS hell, bring the level of hardware polish that PPC hardware has, and totally and completely lead that platform. Apple doesn't stand a chance doing that with 32 bit x86.
If you can accept that Apple wants to move into the Wintel space, this is the kind of rare event that will allow it to happen. If you believe that Apple will do it, you should feel it is likely to happen now and with a deal like this.
um....isnt the target market for the 970 low end servers and desktops? The only reason that I can see that Apple would support the AMD chips is that IBM asked for a consession from Apple to get AltiVec on the 970, and IBM wanted OS X server support for their blade servers. If it were just to sell server software to existing server markets that could potentially cut into hardware sales for Apples, and potentially IBM's servers (if Apple does license IBM to use OS X server, and potentially thin clients for the network).
And even if the Opteron is showing even better performance than a 1.8 ghz 970, the 970 is still a great processor and has already proven (if IBM's press release is something to rely on) capable of even greater speeds. And we are only 12-24 (hopefully) months away from an even greater processor - the 980. Before that we'll also have the choice of the 970+ so the future is not completely dark
FWIW, the 970's die size is a lot smaller (and should be much cheaper) than the Opteron's, for every Opteron rolling off the fab, there should be around 2 970's ...
So ...
Which would you rather have, a single slightly faster Opteron for a fair bit more cash, or dual 970's?
I remembe early in the development in OS X Apple briefly planned for a development environment that could build applications runnable on OS X and on Wintel (I forget what color "box" it was supposed to be). Apple is going to need a tool to write 64 bit apps for the new 970, wouldn't it be spiffy if these apps could very easily run on AMD's 64 bit Wintel offerings. Want to write apps for the new 64 bit platforms? Write it once....on OS X \
I remembe early in the development in OS X Apple briefly planned for a development environment that could build applications runnable on OS X and on Wintel (I forget what color "box" it was supposed to be). Apple is going to need a tool to write 64 bit apps for the new 970, wouldn't it be spiffy if these apps could very easily run on AMD's 64 bit Wintel offerings. Want to write apps for the new 64 bit platforms? Write it once....on OS X \
Rhapsody (Pre OS X/Carbon) had the following compontents that Apple was working on:
Yellow Box- what is now Cocoa
Blue Box- what is now Classic, would run only on Macs using Power PC processors)
Red Box- similar to Classic, this would work only on Macs using the X86 version of Rhapsody. This would run Win 32 compatable programs, basically windows programs. Since it is running on X86, it would not need to emulate the hardware (IBM's OS/2 did this, and some reports had it that OS/2 ran windows software better than Windows did)
Red box would have the potential to kill efforts to move OS X to X86, becouse there would be a tendency for developers to write one code base rather than two for the platform. This is often quoted as one of the reasons OS/2 did not succeed. Also it might not be neccessary if there were enough OS X native programs, becouse these would (Idealy) only need to be recompiled, and they could be distributed in a FAT Bionary (?) form which could be installed on both PowerPC's and X86 platforms, similar to what was done when Apple moved to the PowerPC.
Though I'm skeptical about all the OSX/x86 stuff, it has to be said that both IBM and Apple stand to gain a lot by collaborating with AMD, perhaps even helping to prop them up. It would be a significant blow to IBM and Apple if AMD folded.
So long as the 64-bit x86 market stays fractured, the PPC has a once in a lifetime opportunity here. The enemy of my enemy is my friend and all that....
The point - the POINT of going AMD is not a platform change, but a platform add.
Apple can, though Cocoa (sorry carbon kids), add AMD64 to their PPC line and push out not only to some new markets, but also buy a hedge against a sole CPU supplier.
Provided that Apple can bring their software forward to generate profits, Apple can trim hardware margins to stay competitive and be a first-to-market unix-but-not-unix 64 bit OS vendor.
AMD would be a new market, not a replacement. Mac OS X may become that which NT could never be.
Good God NO! Sure Apple can port the Cocoa API's to x86-64 (or anything else for that matter), but this strategy would confuse the living daylights out of ALL of Apple's developers. Further, it either would DOOM one of the two platforms (probably the PPC line), or it would confuse both so horribly that they would both die off. Companies like Adobe would have to figure out if they were going to support OSX PPC or OSX x86, and the marketshare would only provide enough revenue for one of these platforms.
The only time software vendors have supported two differing binary platforms was during the TRANSITION from 68x to PPC and the reason why the did this was because it was clear where the platform was going. If Apple said that they would release the PPC 601 machines as well as some new 68060 machines, then software vendors would not have readily ported to the PPC platform. In short, you can't release two different platofrms and expect your software vendors to not kill you.
This was to originally run on Windows and OS X. It would have been libraries (or whatever) that would allow for Cocoa apps to run in Win9X or WinNT.
Quote:
Blue Box- what is now Classic, would run only on Macs using Power PC processors)
Correct.
Quote:
Red Box- similar to Classic, this would work only on Macs using the X86 version of Rhapsody.
Not quite. Red Box was actually the rumored Windows emulation environment for PPC (like Virtual PC, but built more like Classic so that it was seemless). I am not sure if Intel OS X ever allowed for running of natvie Windows apps.
Good God NO! Sure Apple can port the Cocoa API's to x86-64 (or anything else for that matter), but this strategy would confuse the living daylights out of ALL of Apple's developers.
Why? It's the same code. You use ProjectBuilder and click the checkbox for PPC and the checkbox for X86-64, hit build, and out pops an application that runs on both platforms. What made Cocoa and the associated technologies so wonderful is that it didn't matter what the hardware was, you just shipped one app that you double-click and run. Hell, the user doesn't even need to know what chip is in there...
Quote:
Further, it either would DOOM one of the two platforms (probably the PPC line), or it would confuse both so horribly that they would both die off. Companies like Adobe would have to figure out if they were going to support OSX PPC or OSX x86, and the marketshare would only provide enough revenue for one of these platforms.
Why? Given where Apple is now, how much would have to change in a 970 based Xserve for an AMD chip to go in? Remember, there's no legacy x86 platform that you need to build to for this - AMD could opt to build around Apple's platform for everything. There may not be any difference. If they both boot off of OpenFirmware, even updates at that level would be supported the same way. Right now, we're using the same GPUs, RAM, USB, Firewire, etc. etc. The differences are really, really small.
And so what if the PPC died off, or the AMD did? So long as Apple is selling xxx number of boxes per year, that's all that matters really. If the boxes were virtually identical except for CPU, I don't understand the support issues except to those developers that need to code to the metal - and there aren't that many of them these days.
Quote:
The only time software vendors have supported two differing binary platforms was during the TRANSITION from 68x to PPC and the reason why the did this was because it was clear where the platform was going. If Apple said that they would release the PPC 601 machines as well as some new 68060 machines, then software vendors would not have readily ported to the PPC platform. In short, you can't release two different platofrms and expect your software vendors to not kill you.
Actually, NeXT supported 4 platforms simultaneously: PPC, x86, Alpha, and Sparc. To the user, it was only one binary platform because the application maintained the code for each platform and invoked it as needed. Put the app on your server and you can double-click it from 4 different hardware platforms and it just works. What killed the concept was that the frameworks hit each platform, but not the OS. The frameworks were then dependent on the whims of each of the OS vendors. That's what killed most of the software vendors, but many of them flocked to the platform because you could write code that ran on 4 platforms. That property is what made WebObjects as desirable as it was - you could write for x86 and when it could no longer drive your site, you could upgrade to a big Sun box - but the software wouldn't have to change at all.
PPC/AMD may not give Apple that kind of scaling, but it would give flexibility by allowing AMD/PPC hardware to mix/match. Another view of this is that 970 might be a transitional product to open up POWER for the high end stuff, leaving AMD for the low-end stuff. After the Xserve runs out of gas, you can drop in a nice 8-way POWER4 box and run the same OS and the same apps. Right now, you need the 970 to bring people along. If developers support the 970 and AMD, and either AMD or the 970 goes away, Apple still has a low end product and they still have POWER to hand enterprise over to. IBM wins both ways, AMD wins both ways, Apple wins both ways.
The Carbon API isn't tied to PowerPC any more than it was tied to 68k. Good thing too considering how much Cocoa apps suck ass. Why you think only Cocoa is platform neutral is beyond me.
There is NO way Apple can pleasantly surprise their developers with a new platform without due notice. If Apple sprung a new platform on us it would be henceforth known as textbook business hara kari. Gawd, just thinking of it makes my skin crawl. Why the hell would I put up with that crap when I can write WinXP apps instead...or RIDE BIKES?
The Carbon API isn't tied to PowerPC any more than it was tied to 68k. Good thing too considering how much Cocoa apps suck ass.
... well, ya got a point there, obviously the latest Cocoa front ended browser, Safari sucks ... oops ... well, hang on, no it doesn't suck, in fact, I'm using it right now ... ok ok, well then, how about any development company that uses Cocoa exclusively, like say, the Omnigroup, their apps must suck right? ...uh, hang on again, they don't suck either, in fact, omnigraffle is getting pretty darn scary good ... geez ... well there's always Apple, a pretty small company that's gone almost exclusively Cocoa for any new software development (minor stuff, like the Finder) ... and you know their software, well, doesn't quite suck either ...
Good thing too considering how much Cocoa apps suck ass.
What the **** are you talkign about? I can find examples of good cocoa apps and bad ones, just like I can find good exampls of carbon apps, and bad ones.
Comments
Originally posted by strobe
Apple would have to give 3rd party developers at least 6 months advance notice of any architecture change. What good is an OS with no apps?
Not to mention all us developers would be POed! Such a move would drive us into WinXP faster than a meteorite landing on cupertino.
Stop passing the hot turd around like it's a box of delicious chocolates.
Um... Er... No.
The point - the POINT of going AMD is not a platform change, but a platform add.
Apple can, though Cocoa (sorry carbon kids), add AMD64 to their PPC line and push out not only to some new markets, but also buy a hedge against a sole CPU supplier.
Provided that Apple can bring their software forward to generate profits, Apple can trim hardware margins to stay competitive and be a first-to-market unix-but-not-unix 64 bit OS vendor.
AMD would be a new market, not a replacement. Mac OS X may become that which NT could never be.
Originally posted by herbivore
The use of AMD chips by Apple would amount to suicide for the company. Unless, of course, it were part of some other strategy to weaken Intel and strengthen IBM.
Apple could do that, and also get into places that don't even look at hardware that isn't x86. Sadly, such places exist. And if it introduces some more confusion into the PC world, so much the better.
I don't think anyone is talking about Apple leaving the PowerPC at this point. Yes, there was talk of that sort during the 500MHz debacle, but with that behind us and a good future mapped out for the next few years, I can't believe Apple has any intention of leaving their baby behind.
The question is, what would Apple release that would be based on an Opteron? I don't really have an answer to that, beyond "a server?" It sure isn't going to be powering iMacs...
Originally posted by johnsonwax
AMD would be a new market, not a replacement. Mac OS X may become that which NT could never be.
And exactly who is that market??? Say that Apple releases a new PM with the Opteron, and even if they release all their iApps, Server software aso, there is almost no applications ready for this market yet. So they essentially would have a very fast and competitive computer with no market. I don't see very many professionals buying macs that cant run any of Adobe's, Macromedia's, Microsoft's or Quark's software for another maybe 12 - 24 months. Unless maybe Apple makes some sort of emulation package available.
And even if the Opteron is showing even better performance than a 1.8 ghz 970, the 970 is still a great processor and has already proven (if IBM's press release is something to rely on) capable of even greater speeds. And we are only 12-24 (hopefully) months away from an even greater processor - the 980. Before that we'll also have the choice of the 970+ so the future is not completely dark
There is also another question. What about portables? Will the Opteron be a good choice for laptops heat and battery-vise? I think not.
and OEMs would sell thoses boxes with OSX as an option, which is boatloads cheaper than windows server in most configs...
power of unix and open source...pretty interface...
Originally posted by atomicham
HA! To hell with you 64-bit wimps, I'm jumping to 640-bit with 4e186 Mbytes of memory. Suckers at Intel will cringe when I have the entire web in RAM.
Originally posted by herbivore
I guess I don't understand the logic of Apple making use of the Opteron or Athlon64.
Neither do I. I don't think Apple is going to switch from PPC anytime soon.
If Apple wanted to make a move over to another architecture, the Itanium would be a better chip.
[Rimmer]
Wrong, wrong. Absolutely brimming over with wrongability. [/Rimmer]
Itanic is very expensive, and despite what you have said elsewhere, Intel can't price it to compete with G4, 970, x86-64, IA-32 or even Alpha. This would be Apple's death blow for sure. X86-64 would be a much better choice (which I don't think will happen - see above).
Apple could port OS X to the Itanium and ride Intel's ability to scale the chip. In my estimation, this would be a much safer route as Intel has shown that it can consistently deliver the results.
The chip will never get cheap compared to PPC without major performance sacrifices which it can't afford since it doesn't have much of a performance lead.
We should be content with IBM's new found committment to the PowerPC. Apple should do everything it can to encourage IBM's ongoing dedication to the chip.
Yes, yes and yes again.
In any case, OS X on the Itanium is within the realm of possibility. OS X on the Opteron/Athlon makes no sense.
[Rimmer] (see above) [/R]
What make far more sense for Apple, if it wan't to get into the high end (which is the only place Itanium is tageted in the current roadmap) is to start using Power chips and develop 4x, 8x and up servers, etc. Or port OS X to IBM's hardware and sell them together in partnership with IBM.
IBM is going to stick with PPC for a long time (as you imply). They will be commited to it in their blades and workstations (as a 604 and Power3 replacement) and will make new, faster ones. Power4 was the first step in IBMs refocusing on the fact that it had all the technology to develop the best processors but was futsing around with a team than wasn't given enough attention from the top. Now it has that attention. IBM is talking about the death of all 64-bitters except Itanic and Power. They even want to squash Itanic. They will stay in the game and give Apple what they need.
Who knows? Apple's market share might rise enough for Mot to get really committed to desktop and workstation design and we could see Intel and AMD in the PC world with IBM and Mot in the PPC world, all strong and healthily competing and giving us good products.
MM
Originally posted by keyboardf12
I thought all this Opteron talk was more along the lines of MacOSX server being recompiled for it...
and OEMs would sell thoses boxes with OSX as an option, which is boatloads cheaper than windows server in most configs...
power of unix and open source...pretty interface...
Yes.
but not until this fall when Panther Server comes out (at the earliest)
Originally posted by NETROMac
And exactly who is that market??? Say that Apple releases a new PM with the Opteron, and even if they release all their iApps, Server software aso, there is almost no applications ready for this market yet. So they essentially would have a very fast and competitive computer with no market. I don't see very many professionals buying macs that cant run any of Adobe's, Macromedia's, Microsoft's or Quark's software for another maybe 12 - 24 months. Unless maybe Apple makes some sort of emulation package available.
And even if the Opteron is showing even better performance than a 1.8 ghz 970, the 970 is still a great processor and has already proven (if IBM's press release is something to rely on) capable of even greater speeds. And we are only 12-24 (hopefully) months away from an even greater processor - the 980. Before that we'll also have the choice of the 970+ so the future is not completely dark
Well, Opteron is mainly targetted at servers. The market is those who would just automatically grab an x86 solution so they have unix/linux/Win compatibility. Apple could potentially build AMD powered Xserves, clusters, grid setups. Companies can confidently buy them with the knowledge that if OS X doesn't work out, they can fall back to traditional solutions. Apple doesn't need to deliver iApps here for a while. Only developers targetting servers need to jump immediately, and many of them deal with POSIX or Cocoa APIs, so that's probably a snap for them.
My guess is that Apple could bring products to bear on a new platform faster than anyone else including MS since Cocoa is designed to target multiple platforms. Anyone currently writing cocoa apps are likely not much more than a recompile away.
Since AMD/OS X could be 64bit only, it doesn't uncork the genie implied by x86 support, Apple could maintain their hardware control and reach into enterprise that want their MS backup. It's the fact that Opt and 970 are close in performance that makes this attractive for Apple since customers don't have to chase benchmarks to decide. It gives Apple a backup in case one or the other craps out. It gives developers a new family of hardware to go after, which is *good* since Apple's dev tools make it so freaking easy to support both, and it gives Apple the ability to lead on all leading 64bit platforms.
If Apple does want to go after a broader market and they want to go into MSs turf (compete head-to-head with Windows on the same hardware), they need some kind of pristine environment like this to jump into that doesn't suffer from legacy driver problems, and all of that. Apple could bring OpenFirmware to Opteron, do away with BIOS hell, bring the level of hardware polish that PPC hardware has, and totally and completely lead that platform. Apple doesn't stand a chance doing that with 32 bit x86.
If you can accept that Apple wants to move into the Wintel space, this is the kind of rare event that will allow it to happen. If you believe that Apple will do it, you should feel it is likely to happen now and with a deal like this.
Originally posted by NETROMac
And even if the Opteron is showing even better performance than a 1.8 ghz 970, the 970 is still a great processor and has already proven (if IBM's press release is something to rely on) capable of even greater speeds. And we are only 12-24 (hopefully) months away from an even greater processor - the 980. Before that we'll also have the choice of the 970+ so the future is not completely dark
FWIW, the 970's die size is a lot smaller (and should be much cheaper) than the Opteron's, for every Opteron rolling off the fab, there should be around 2 970's ...
So ...
Which would you rather have, a single slightly faster Opteron for a fair bit more cash, or dual 970's?
Hmmmmmmm?
Originally posted by BobtheTomato
I remembe early in the development in OS X Apple briefly planned for a development environment that could build applications runnable on OS X and on Wintel (I forget what color "box" it was supposed to be). Apple is going to need a tool to write 64 bit apps for the new 970, wouldn't it be spiffy if these apps could very easily run on AMD's 64 bit Wintel offerings. Want to write apps for the new 64 bit platforms? Write it once....on OS X
Rhapsody (Pre OS X/Carbon) had the following compontents that Apple was working on:
Yellow Box- what is now Cocoa
Blue Box- what is now Classic, would run only on Macs using Power PC processors)
Red Box- similar to Classic, this would work only on Macs using the X86 version of Rhapsody. This would run Win 32 compatable programs, basically windows programs. Since it is running on X86, it would not need to emulate the hardware (IBM's OS/2 did this, and some reports had it that OS/2 ran windows software better than Windows did)
Red box would have the potential to kill efforts to move OS X to X86, becouse there would be a tendency for developers to write one code base rather than two for the platform. This is often quoted as one of the reasons OS/2 did not succeed. Also it might not be neccessary if there were enough OS X native programs, becouse these would (Idealy) only need to be recompiled, and they could be distributed in a FAT Bionary (?) form which could be installed on both PowerPC's and X86 platforms, similar to what was done when Apple moved to the PowerPC.
So long as the 64-bit x86 market stays fractured, the PPC has a once in a lifetime opportunity here. The enemy of my enemy is my friend and all that....
Originally posted by johnsonwax
Um... Er... No.
The point - the POINT of going AMD is not a platform change, but a platform add.
Apple can, though Cocoa (sorry carbon kids), add AMD64 to their PPC line and push out not only to some new markets, but also buy a hedge against a sole CPU supplier.
Provided that Apple can bring their software forward to generate profits, Apple can trim hardware margins to stay competitive and be a first-to-market unix-but-not-unix 64 bit OS vendor.
AMD would be a new market, not a replacement. Mac OS X may become that which NT could never be.
Good God NO! Sure Apple can port the Cocoa API's to x86-64 (or anything else for that matter), but this strategy would confuse the living daylights out of ALL of Apple's developers. Further, it either would DOOM one of the two platforms (probably the PPC line), or it would confuse both so horribly that they would both die off. Companies like Adobe would have to figure out if they were going to support OSX PPC or OSX x86, and the marketshare would only provide enough revenue for one of these platforms.
The only time software vendors have supported two differing binary platforms was during the TRANSITION from 68x to PPC and the reason why the did this was because it was clear where the platform was going. If Apple said that they would release the PPC 601 machines as well as some new 68060 machines, then software vendors would not have readily ported to the PPC platform. In short, you can't release two different platofrms and expect your software vendors to not kill you.
Yellow Box- what is now Cocoa
This was to originally run on Windows and OS X. It would have been libraries (or whatever) that would allow for Cocoa apps to run in Win9X or WinNT.
Blue Box- what is now Classic, would run only on Macs using Power PC processors)
Correct.
Red Box- similar to Classic, this would work only on Macs using the X86 version of Rhapsody.
Not quite. Red Box was actually the rumored Windows emulation environment for PPC (like Virtual PC, but built more like Classic so that it was seemless). I am not sure if Intel OS X ever allowed for running of natvie Windows apps.
Originally posted by MacAficionado
I found interesting also that AMD plans to release the AMD64 in September, around the time Apple usually upgrades hardware.
I'm not saying that they will go x86, but that they are difinitely cooperating. The axis of goodness: IBM, AMD, Apple
Can't wait to find out!
Yes, but it's only really going to get "confirmed" if we can catagorically state that AMD will release the AMD64 on a Tuesday.
Until then this is just fluff.
Originally posted by Yevgeny
Good God NO! Sure Apple can port the Cocoa API's to x86-64 (or anything else for that matter), but this strategy would confuse the living daylights out of ALL of Apple's developers.
Why? It's the same code. You use ProjectBuilder and click the checkbox for PPC and the checkbox for X86-64, hit build, and out pops an application that runs on both platforms. What made Cocoa and the associated technologies so wonderful is that it didn't matter what the hardware was, you just shipped one app that you double-click and run. Hell, the user doesn't even need to know what chip is in there...
Further, it either would DOOM one of the two platforms (probably the PPC line), or it would confuse both so horribly that they would both die off. Companies like Adobe would have to figure out if they were going to support OSX PPC or OSX x86, and the marketshare would only provide enough revenue for one of these platforms.
Why? Given where Apple is now, how much would have to change in a 970 based Xserve for an AMD chip to go in? Remember, there's no legacy x86 platform that you need to build to for this - AMD could opt to build around Apple's platform for everything. There may not be any difference. If they both boot off of OpenFirmware, even updates at that level would be supported the same way. Right now, we're using the same GPUs, RAM, USB, Firewire, etc. etc. The differences are really, really small.
And so what if the PPC died off, or the AMD did? So long as Apple is selling xxx number of boxes per year, that's all that matters really. If the boxes were virtually identical except for CPU, I don't understand the support issues except to those developers that need to code to the metal - and there aren't that many of them these days.
The only time software vendors have supported two differing binary platforms was during the TRANSITION from 68x to PPC and the reason why the did this was because it was clear where the platform was going. If Apple said that they would release the PPC 601 machines as well as some new 68060 machines, then software vendors would not have readily ported to the PPC platform. In short, you can't release two different platofrms and expect your software vendors to not kill you.
Actually, NeXT supported 4 platforms simultaneously: PPC, x86, Alpha, and Sparc. To the user, it was only one binary platform because the application maintained the code for each platform and invoked it as needed. Put the app on your server and you can double-click it from 4 different hardware platforms and it just works. What killed the concept was that the frameworks hit each platform, but not the OS. The frameworks were then dependent on the whims of each of the OS vendors. That's what killed most of the software vendors, but many of them flocked to the platform because you could write code that ran on 4 platforms. That property is what made WebObjects as desirable as it was - you could write for x86 and when it could no longer drive your site, you could upgrade to a big Sun box - but the software wouldn't have to change at all.
PPC/AMD may not give Apple that kind of scaling, but it would give flexibility by allowing AMD/PPC hardware to mix/match. Another view of this is that 970 might be a transitional product to open up POWER for the high end stuff, leaving AMD for the low-end stuff. After the Xserve runs out of gas, you can drop in a nice 8-way POWER4 box and run the same OS and the same apps. Right now, you need the 970 to bring people along. If developers support the 970 and AMD, and either AMD or the 970 goes away, Apple still has a low end product and they still have POWER to hand enterprise over to. IBM wins both ways, AMD wins both ways, Apple wins both ways.
There is NO way Apple can pleasantly surprise their developers with a new platform without due notice. If Apple sprung a new platform on us it would be henceforth known as textbook business hara kari. Gawd, just thinking of it makes my skin crawl. Why the hell would I put up with that crap when I can write WinXP apps instead...or RIDE BIKES?
Originally posted by strobe
The Carbon API isn't tied to PowerPC any more than it was tied to 68k. Good thing too considering how much Cocoa apps suck ass.
... well, ya got a point there, obviously the latest Cocoa front ended browser, Safari sucks ... oops ... well, hang on, no it doesn't suck, in fact, I'm using it right now ... ok ok, well then, how about any development company that uses Cocoa exclusively, like say, the Omnigroup, their apps must suck right? ...uh, hang on again, they don't suck either, in fact, omnigraffle is getting pretty darn scary good ... geez ... well there's always Apple, a pretty small company that's gone almost exclusively Cocoa for any new software development (minor stuff, like the Finder) ... and you know their software, well, doesn't quite suck either ...
Hmmmmm do we see a pattern here?
Originally posted by strobe
Good thing too considering how much Cocoa apps suck ass.
What the **** are you talkign about? I can find examples of good cocoa apps and bad ones, just like I can find good exampls of carbon apps, and bad ones.
Mind elaborating as to why "Cocoa apps suck ass"?