Ding Ding! Give that man from Chapel Hill a cookie. This opens the door to a lot more flexibility in Apple's future. Unlike NeXT, they're not dumping the hardware with this change, but it offers them to be much more agile with what they do under the hood of any of their future hardware products. I mean, what if they do want to use a PPC chip in the future? This doesn't proclude such a thing at all.
I think the most interesting aspect is that it's really going to push a lot of developers into the higher-level APIs, XCode and Cocoa that Apple wants them to use anyway. Once that happens, if it hasn't already in some cases, Apple can go beyond even Intel and PPC offerings if any alternatives appear on the horizon.
Oh, and for eveyone who's decided that having an Intel chip inside a Mac someday means you might as well buy a Dell next, congratulations, you've apparently completely missed out on the Macintosh experience.
PowerPC isn't dead by any means. The real problem has been with the portables, but the high end of the PowerPC is promising.... we may never see a complete transition to Intel and be surprised by yet another about face. Jobs is asking developers to ready Fat Binaries all around, which would make a reverse move easier.... In the unlikely event that Apple and IBM partner again, Apple will be ready.
Or maybe this is just all wishful thinking ....
Until yesterday we speculated if and how Intel will enter the game. Now we speculate if and how IBM will enter again.
That's the deal. Jobs declared the PowerPC dead. No backdoor open. But now who will trust him again?
I think that this isn't really a switch to Intel - it is a swtich to "processor independance".
We will have fat binaries for the rest of time, and Apple will use whichever processor is best at a particular time.
G6 towers, x86 laptops, and maybe later something else.
That's nice to dream about, but the official Apple position is to drop the PowerPC processors completely from the beginning of 2008. Once they do that, we will have those fat binaries until the x86 machines hit a critical mass. After that expect them to drop software support for old PPC machines.
I am really surprised by the reactionary, knee-jerk responses on this forum. Most of the posts here from "developers" are just as irrational as those from laymen. Yes, there are many unanswered questions right now, but that doesn't give your assumptions (i.e. user hurdles, loss of 64-bit, etc) any credibility. In fact, if you are a real developer, then why aren't you at WWDC? I'm sure there will be many more answers provided there.
For over ten years, I have waited for the PPC to deliver...I'm still waiting. Are your memories so short that you've forgotten all the drama that has surrounded Apple's CPUs?
Forget about elegant chip architectures, this is business 101. Did you notice IBM's reaction to the loss of Apple's business? Neither did most people, because there was none. Apple's switch does nothing to impact IBM's bottom. If the company does not care about loosing your business, then how can you expect them to work to keep it?
These are simple economies of scale. Apple does not sell enough computers to garner the undivided attention of a mainstream chip manufacturer. Yet, for years they have tried. Motorola wanted to make phones and IBM wants to make its own servers (high margins) and game consoles (high volume), both firms treated Apple as an after thought. How can the Mac market-share grow in these conditions? It can't.
This is a great day for the future of the Mac platform. Finally, we have a chip maker that is dedicated to making great chips that are competitive in the consumer, workstation, and server space. Apple is no longer the ugly step-child, but a valued customer of a key business area.
Although I think it the right decision I am actually somewhat concerned for being on the same hardware. Whereas before it was possible to blame the hardware for the disparity, or the lack of optimisation it will now be possible to outfit identical hardware almost and compare the performance of the OS alone, which may highlight a few shortcomings. For all OS Xs polish and multitasking it hasn't had the sheer speed in a lot of respects that Windows does.
Yes, but now apple will have to make it feel as fast, if not faster than windows. Otherwise, who would buy it.
Yes, but now apple will have to make it feel as fast, if not faster than windows. Otherwise, who would buy it.
How can you tell the difference? Really - my mac mini feels lightning quick, and I can't tell the difference between it and one of the G5 towers when doing normal work.
Processor speed is not that interesting anymore - what is interesting is low-cost. If all the cars are 1000+ hp, who cares if you have 1000 hp or 10,000 hp.
I've had some time to chew on this now and here are my thoughts (not that anyone cares, but consider it therapy).
1. The importance of laptops. There were numbers out recently that laptop sales have surpassed desktop sales for the first time. Apple's laptop line-up is underpowered. IBM and Freescale don't seem to be able/willing to provide the low power/high performance chips Apple needs to keep their laptops from slipping further down the performance slope.
2. Apple is small potatoes to IBM and Freescale. I read in the WSJ that Apple sales are 2-3% of IBM and Freescale's chip revenue (and chip revenue is a small percentage of IBM's total revenue). There was a quote in the WSJ this morning from IBM saying essentially, "we're following the money". Apple's CPU business does not represent enough money for IBM or Freescale to care about keeping Apple as a customer.
3. Intel makes and sells LOTS of PC CPUs for both laptop and desktop machines. The kinds of CPUs that Apple needs - and lots of other PC makers (Dell, HP, et al.). What's good CPU-wise for Wintel box makers is good for Apple, and Intel is interested in keeping the Wintel world happy, thus Apple gets some joy as a convenient by-product.
4. This may free up Intel to do some interesting things chip-wise. I wonder if Microsoft has influenced Intel to keep their ISA ugly to make life easier for MS? If Apple is a viable (not major, just viable) CPU customer, perhaps Intel can release some cool stuff they've been wanting to do and show that they are not MS's pawn?
I believe all those factors make the Intel "switch" a good/only choice. As has been pointed out, there still is room for monster PPC workstations - which seems to be the realm where the PPC shines. But the laptop line goes to Intel. For the "average" PC user, they won't notice the difference or care what CPU is inside as long as fat/universal binaries and Rosetta work as advertised.
Things get more interesting when you start talking Motion, FCP, GarageBand, and anything that is heavily AltiVec optimized/reliant or benefits from the PPC's awesome bus speeds. At the high-end there may definitely end up being clear performance reasons to choose a PPC or Intel based Mac. While Mathematica may run on Intel, it may be a lot faster on PPC; while Adobe CS may run on PPC it may be faster on Intel. I don't know how that will work out for Apple. Will they be able to crow, "choose the chip that's right for you!" or end up fragmenting the market, causing confusion and frustration, and ultimately lose out to the Windows and Linux camps?
I think this will work out alright for Apple. If the OS and development platform they have has been abstracted away from the CPU hardware enough to support this sort of chip-agnosticism, they should be able to migrate to different and even more powerful CPUs in the future with little trouble.
I couldn't believe that this would hsppen. But after watching the kenote I realise that it's what Apple wants to happen.
Sit down at your iMac or whatever and just use it. If everything works the way it's supposed to, what's the difference.
No one even thought about it until he said it was running on Intel.
Do most people care if the engine block in their car is aluminum or iron? Not really.
Apple seems to do what other companies have problems doing. But Apple isn't the only company thast's done this. Sun and SGI have as well. so have others.
Yes, and Sun's not doing so well, and oh, how about that SGI? I've never seen a company that could dependably choose the WORST option each time, every time, like SGI...
Things get more interesting when you start talking Motion, FCP, GarageBand, and anything that is heavily AltiVec optimized/reliant or benefits from the PPC's awesome bus speeds. At the high-end there may definitely end up being clear performance reasons to choose a PPC or Intel based Mac. While Mathematica may run on Intel, it may be a lot faster on PPC; while Adobe CS may run on PPC it may be faster on Intel. I don't know how that will work out for Apple. Will they be able to crow, "choose the chip that's right for you!" or end up fragmenting the market, causing confusion and frustration, and ultimately lose out to the Windows and Linux camps?
That is why they are starting out with the low-end macs, because the current PowerPC towers would be lower performance if they switched to x86 now.
Yes, but now apple will have to make it feel as fast, if not faster than windows. Otherwise, who would buy it.
It will be very interesting, and I bet the first thing everyone's going to want to know is how does [insert your favorite intense application] on a 3.6Ghz Intel-based Mac compare to Photoshop on an 3.6Ghz Wintel box, natively. I'm guessing first results will show things running slower on Intel-based OS X. Then we'll all say, "yeah, but wait until it's optimized!". We'll surely lag behind for a few years.
Here's my guess: a fresh-out-of-the-box comparison of a OS X and Windows running the same Intel chip will have the Windows applications coming out ahead. But a comparison of applications running on an OS X and Windows box that have been in use for six months will have the OS X applications come out ahead, due to all the spyware and garbage bogging down the Windows-based Intel box.
1. Jobs used Intel processors in the boxes he built at NeXT, the computer company he founded...It had Intel inside.
Actually, this is not true. NEXT never produced Intel-based machines (at least not publicly). They began with the Motorola 68030 (some say the first NeXT computer was a the next Mac computer). Then the 68040. There were rumors of a Motorola 88000 or PowerPC based NeXT computer before the hardware went away.
NeXT did however make NEXTSTEP publicly available for the following processors:
It will be very interesting, and I bet the first thing everyone's going to want to know is how does [insert your favorite intense application] on a 3.6Ghz Intel-based Mac compare to Photoshop on an 3.6Ghz Wintel box, natively. I'm guessing first results will show things running slower on Intel-based OS X. Then we'll all say, "yeah, but wait until it's optimized!". We'll surely lag behind for a few years.
Here's my guess: a fresh-out-of-the-box comparison of a OS X and Windows running the same Intel chip will have the Windows applications coming out ahead. But a comparison of applications running on an OS X and Windows box that have been in use for six months will have the OS X applications come out ahead, due to all the spyware and garbage bogging down the Windows-based Intel box.
I do agree with you, but what Apple have going for them is that they will have programed OS X for their hardware, so hopefully they will be able to optomize the software and hardware IMEDIATELY.
I predict the Apple speed will be greater over a Wintel box speed. The reason I say this is that my G4 Cube is running 450, yes some it is slower than todays modern hardware, but if you care my cube to an AMD anthlon running at 1.2 to 1.5, I would have to say that my cube runs better. The problem with Windows other than the sercurity problems is that they require so many resources just to run the operating system. I have found that Apple's OS is extremely effiecent and it use the resources it has at hand wisely.
a couple more thoughts on the forthcoming change in processor suppliers at Apple:
1. Jobs used Intel processors in the boxes he built at NeXT, the computer company he founded after he was kicked out of Apple (OK, he lost a palace revolution) in the mid-80s. Some say he was replaced by John Sculley, which is not so, since Jobs was not at the time the CEO.
However, return the mind to NeXT. It had Intel inside.
Really? Mine's a 68040...you'd better hang on to yours, it'll be quite a one-of-a-kind collector's item...
It is not processor independence, exactly. Yes, OS X runs on both processors but Apple is phasing out of PPC and into Intel. If I had an intel iMac and intel iBook (see my sig) I would not be worried.
If it was processor independence and he had said, "Don't worry, you can pick your processor (intel or PPC) by the end of 2007 and beyond," I would be cool as a refrigerated mustard and would be estatic. He did not say that. It is obvious he is expecting heat just from the way he says things. You can tell he is really watching his step so current sales don't fall off.
So basically, in 10 more years he can pull out another double life PPC Keynote and say guess what, Intel Sucks, look what IBM is doing, our OS runs on it, and here we go again. I guess it is just a mental thing but that is aggravating.
"We know transitions." That can be good and can be bad. Customers, like me, only want to put up with so much change because it becomes tiresome and makes me think that software I am using today won't be used on current hardware 4 years from now. Even though I will most likely have new stuff by then. See how silly some consumers are.
I predict the Apple speed will be greater over a Wintel box speed. The reason I say this is that my G4 Cube is running 450, yes some it is slower than todays modern hardware, but if you care my cube to an AMD anthlon running at 1.2 to 1.5, I would have to say that my cube runs better.
Guess what, the G4 has the Altivec unit and OS X makes much use of it. The x86 processors don't have that. Ooops.
This is a great day for the future of the Mac platform. Finally, we have a chip maker that is dedicated to making great chips that are competitive in the consumer, workstation, and server space. Apple is no longer the ugly step-child, but a valued customer of a key business area.
Let the FUD go!
Sorry, look at how many chips Apple will buy vs. how many Dell will buy. Apple is still the ugly step child. They are kind of a trophy for Intel--something they can brag about finally capturing--but Apple will make no dent in their bottom line. And if they start to hurt Dell, and Dell threatens to go to AMD? Well, we might find those nice Intel processors come a little slower.
Let go of the RFD and I'll let go of the FUD. Then maybe we can discuss the unanswered questions this leaves, and the dangers and possibilities it represents.
Comments
We will have fat binaries for the rest of time, and Apple will use whichever processor is best at a particular time.
G6 towers, x86 laptops, and maybe later something else.
I think the most interesting aspect is that it's really going to push a lot of developers into the higher-level APIs, XCode and Cocoa that Apple wants them to use anyway. Once that happens, if it hasn't already in some cases, Apple can go beyond even Intel and PPC offerings if any alternatives appear on the horizon.
Oh, and for eveyone who's decided that having an Intel chip inside a Mac someday means you might as well buy a Dell next, congratulations, you've apparently completely missed out on the Macintosh experience.
Originally posted by Sybaritic
PowerPC isn't dead by any means. The real problem has been with the portables, but the high end of the PowerPC is promising.... we may never see a complete transition to Intel and be surprised by yet another about face. Jobs is asking developers to ready Fat Binaries all around, which would make a reverse move easier.... In the unlikely event that Apple and IBM partner again, Apple will be ready.
Or maybe this is just all wishful thinking ....
Until yesterday we speculated if and how Intel will enter the game. Now we speculate if and how IBM will enter again.
That's the deal. Jobs declared the PowerPC dead. No backdoor open. But now who will trust him again?
Originally posted by e1618978
I think that this isn't really a switch to Intel - it is a swtich to "processor independance".
We will have fat binaries for the rest of time, and Apple will use whichever processor is best at a particular time.
G6 towers, x86 laptops, and maybe later something else.
That's nice to dream about, but the official Apple position is to drop the PowerPC processors completely from the beginning of 2008. Once they do that, we will have those fat binaries until the x86 machines hit a critical mass. After that expect them to drop software support for old PPC machines.
For over ten years, I have waited for the PPC to deliver...I'm still waiting. Are your memories so short that you've forgotten all the drama that has surrounded Apple's CPUs?
Forget about elegant chip architectures, this is business 101. Did you notice IBM's reaction to the loss of Apple's business? Neither did most people, because there was none. Apple's switch does nothing to impact IBM's bottom. If the company does not care about loosing your business, then how can you expect them to work to keep it?
These are simple economies of scale. Apple does not sell enough computers to garner the undivided attention of a mainstream chip manufacturer. Yet, for years they have tried. Motorola wanted to make phones and IBM wants to make its own servers (high margins) and game consoles (high volume), both firms treated Apple as an after thought. How can the Mac market-share grow in these conditions? It can't.
This is a great day for the future of the Mac platform. Finally, we have a chip maker that is dedicated to making great chips that are competitive in the consumer, workstation, and server space. Apple is no longer the ugly step-child, but a valued customer of a key business area.
Let the FUD go!
Originally posted by Telomar
Just to raise something from the first page.
Although I think it the right decision I am actually somewhat concerned for being on the same hardware. Whereas before it was possible to blame the hardware for the disparity, or the lack of optimisation it will now be possible to outfit identical hardware almost and compare the performance of the OS alone, which may highlight a few shortcomings. For all OS Xs polish and multitasking it hasn't had the sheer speed in a lot of respects that Windows does.
Yes, but now apple will have to make it feel as fast, if not faster than windows. Otherwise, who would buy it.
Originally posted by anand
Yes, but now apple will have to make it feel as fast, if not faster than windows. Otherwise, who would buy it.
How can you tell the difference? Really - my mac mini feels lightning quick, and I can't tell the difference between it and one of the G5 towers when doing normal work.
Processor speed is not that interesting anymore - what is interesting is low-cost. If all the cars are 1000+ hp, who cares if you have 1000 hp or 10,000 hp.
1. The importance of laptops. There were numbers out recently that laptop sales have surpassed desktop sales for the first time. Apple's laptop line-up is underpowered. IBM and Freescale don't seem to be able/willing to provide the low power/high performance chips Apple needs to keep their laptops from slipping further down the performance slope.
2. Apple is small potatoes to IBM and Freescale. I read in the WSJ that Apple sales are 2-3% of IBM and Freescale's chip revenue (and chip revenue is a small percentage of IBM's total revenue). There was a quote in the WSJ this morning from IBM saying essentially, "we're following the money". Apple's CPU business does not represent enough money for IBM or Freescale to care about keeping Apple as a customer.
3. Intel makes and sells LOTS of PC CPUs for both laptop and desktop machines. The kinds of CPUs that Apple needs - and lots of other PC makers (Dell, HP, et al.). What's good CPU-wise for Wintel box makers is good for Apple, and Intel is interested in keeping the Wintel world happy, thus Apple gets some joy as a convenient by-product.
4. This may free up Intel to do some interesting things chip-wise. I wonder if Microsoft has influenced Intel to keep their ISA ugly to make life easier for MS? If Apple is a viable (not major, just viable) CPU customer, perhaps Intel can release some cool stuff they've been wanting to do and show that they are not MS's pawn?
I believe all those factors make the Intel "switch" a good/only choice. As has been pointed out, there still is room for monster PPC workstations - which seems to be the realm where the PPC shines. But the laptop line goes to Intel. For the "average" PC user, they won't notice the difference or care what CPU is inside as long as fat/universal binaries and Rosetta work as advertised.
Things get more interesting when you start talking Motion, FCP, GarageBand, and anything that is heavily AltiVec optimized/reliant or benefits from the PPC's awesome bus speeds. At the high-end there may definitely end up being clear performance reasons to choose a PPC or Intel based Mac. While Mathematica may run on Intel, it may be a lot faster on PPC; while Adobe CS may run on PPC it may be faster on Intel. I don't know how that will work out for Apple. Will they be able to crow, "choose the chip that's right for you!" or end up fragmenting the market, causing confusion and frustration, and ultimately lose out to the Windows and Linux camps?
I think this will work out alright for Apple. If the OS and development platform they have has been abstracted away from the CPU hardware enough to support this sort of chip-agnosticism, they should be able to migrate to different and even more powerful CPUs in the future with little trouble.
- Jasen.
Originally posted by melgross
That's the way it should be.
I couldn't believe that this would hsppen. But after watching the kenote I realise that it's what Apple wants to happen.
Sit down at your iMac or whatever and just use it. If everything works the way it's supposed to, what's the difference.
No one even thought about it until he said it was running on Intel.
Do most people care if the engine block in their car is aluminum or iron? Not really.
Apple seems to do what other companies have problems doing. But Apple isn't the only company thast's done this. Sun and SGI have as well. so have others.
Yes, and Sun's not doing so well, and oh, how about that SGI? I've never seen a company that could dependably choose the WORST option each time, every time, like SGI...
Things get more interesting when you start talking Motion, FCP, GarageBand, and anything that is heavily AltiVec optimized/reliant or benefits from the PPC's awesome bus speeds. At the high-end there may definitely end up being clear performance reasons to choose a PPC or Intel based Mac. While Mathematica may run on Intel, it may be a lot faster on PPC; while Adobe CS may run on PPC it may be faster on Intel. I don't know how that will work out for Apple. Will they be able to crow, "choose the chip that's right for you!" or end up fragmenting the market, causing confusion and frustration, and ultimately lose out to the Windows and Linux camps?
That is why they are starting out with the low-end macs, because the current PowerPC towers would be lower performance if they switched to x86 now.
e1618978 and bozza have just made one of the most intelligent, rational and realistically hopeful statements on this thread...
Originally posted by anand
Yes, but now apple will have to make it feel as fast, if not faster than windows. Otherwise, who would buy it.
It will be very interesting, and I bet the first thing everyone's going to want to know is how does [insert your favorite intense application] on a 3.6Ghz Intel-based Mac compare to Photoshop on an 3.6Ghz Wintel box, natively. I'm guessing first results will show things running slower on Intel-based OS X. Then we'll all say, "yeah, but wait until it's optimized!". We'll surely lag behind for a few years.
Here's my guess: a fresh-out-of-the-box comparison of a OS X and Windows running the same Intel chip will have the Windows applications coming out ahead. But a comparison of applications running on an OS X and Windows box that have been in use for six months will have the OS X applications come out ahead, due to all the spyware and garbage bogging down the Windows-based Intel box.
Originally posted by bozza
1. Jobs used Intel processors in the boxes he built at NeXT, the computer company he founded...It had Intel inside.
Actually, this is not true. NEXT never produced Intel-based machines (at least not publicly). They began with the Motorola 68030 (some say the first NeXT computer was a the next Mac computer). Then the 68040. There were rumors of a Motorola 88000 or PowerPC based NeXT computer before the hardware went away.
NeXT did however make NEXTSTEP publicly available for the following processors:
1. 68K (their own machines)
2. Hewlett-Packard PA-RISC machines
3. SPARC
4. Intel
Originally posted by e1618978
I think that this isn't really a switch to Intel - it is a swtich to "processor independance".
We will have fat binaries for the rest of time, and Apple will use whichever processor is best at a particular time.
G6 towers, x86 laptops, and maybe later something else.
I think this is exactly correct.
Originally posted by Cory Bauer
It will be very interesting, and I bet the first thing everyone's going to want to know is how does [insert your favorite intense application] on a 3.6Ghz Intel-based Mac compare to Photoshop on an 3.6Ghz Wintel box, natively. I'm guessing first results will show things running slower on Intel-based OS X. Then we'll all say, "yeah, but wait until it's optimized!". We'll surely lag behind for a few years.
Here's my guess: a fresh-out-of-the-box comparison of a OS X and Windows running the same Intel chip will have the Windows applications coming out ahead. But a comparison of applications running on an OS X and Windows box that have been in use for six months will have the OS X applications come out ahead, due to all the spyware and garbage bogging down the Windows-based Intel box.
I do agree with you, but what Apple have going for them is that they will have programed OS X for their hardware, so hopefully they will be able to optomize the software and hardware IMEDIATELY.
I predict the Apple speed will be greater over a Wintel box speed. The reason I say this is that my G4 Cube is running 450, yes some it is slower than todays modern hardware, but if you care my cube to an AMD anthlon running at 1.2 to 1.5, I would have to say that my cube runs better. The problem with Windows other than the sercurity problems is that they require so many resources just to run the operating system. I have found that Apple's OS is extremely effiecent and it use the resources it has at hand wisely.
Originally posted by bozza
a couple more thoughts on the forthcoming change in processor suppliers at Apple:
1. Jobs used Intel processors in the boxes he built at NeXT, the computer company he founded after he was kicked out of Apple (OK, he lost a palace revolution) in the mid-80s. Some say he was replaced by John Sculley, which is not so, since Jobs was not at the time the CEO.
However, return the mind to NeXT. It had Intel inside.
Really? Mine's a 68040...you'd better hang on to yours, it'll be quite a one-of-a-kind collector's item...
If it was processor independence and he had said, "Don't worry, you can pick your processor (intel or PPC) by the end of 2007 and beyond," I would be cool as a refrigerated mustard and would be estatic. He did not say that. It is obvious he is expecting heat just from the way he says things. You can tell he is really watching his step so current sales don't fall off.
So basically, in 10 more years he can pull out another double life PPC Keynote and say guess what, Intel Sucks, look what IBM is doing, our OS runs on it, and here we go again. I guess it is just a mental thing but that is aggravating.
"We know transitions." That can be good and can be bad. Customers, like me, only want to put up with so much change because it becomes tiresome and makes me think that software I am using today won't be used on current hardware 4 years from now. Even though I will most likely have new stuff by then. See how silly some consumers are.
Originally posted by Keda
These are simple economies of scale. Apple does not sell enough computers to garner the undivided attention of a mainstream chip manufacturer.
Ummm...so why go to Intel, who's even more mainstream than IBM?
Maybe this could have been MIPS big comeback....
Originally posted by iGrant
I predict the Apple speed will be greater over a Wintel box speed. The reason I say this is that my G4 Cube is running 450, yes some it is slower than todays modern hardware, but if you care my cube to an AMD anthlon running at 1.2 to 1.5, I would have to say that my cube runs better.
Guess what, the G4 has the Altivec unit and OS X makes much use of it. The x86 processors don't have that. Ooops.
Originally posted by Keda
This is a great day for the future of the Mac platform. Finally, we have a chip maker that is dedicated to making great chips that are competitive in the consumer, workstation, and server space. Apple is no longer the ugly step-child, but a valued customer of a key business area.
Let the FUD go!
Sorry, look at how many chips Apple will buy vs. how many Dell will buy. Apple is still the ugly step child. They are kind of a trophy for Intel--something they can brag about finally capturing--but Apple will make no dent in their bottom line. And if they start to hurt Dell, and Dell threatens to go to AMD? Well, we might find those nice Intel processors come a little slower.
Let go of the RFD and I'll let go of the FUD. Then maybe we can discuss the unanswered questions this leaves, and the dangers and possibilities it represents.