the daily snivel
Hot Apple on Intel Action.
Everyone else is declaring that "hell has frozen over" (literally: google the phrase "Hell freezes over" and the word Apple and see how many hits you get) due to the fact that Apple has announce it is moving its computers to an Intel processor-based platform beginning in 2006. I wouldn't go so far as to say that the devil is looking for his red longjohns -- I'd say it was probably the best decision given the sad reality that IBM and the PowerPC chip, while promising much, rarely deliver at the end of the day.
For years, of course, Apple's computers were based on the Motorola 68000 series processor, followed by the PowerPC design. Most recently, Apple began buying its coveted G5 PowerPC processors from IBM, but has been stung by IBMs ability to provide these processors in great volume. When Steve Jobs introduced the G5 processor in the first 2 gigahertz Power Macs over a year ago, he proudly stated that the G5 would have a clock speed of 3 gigahertz within twelve months. This never materialized, and Apple's top of the line Power Mac G5 has two 2.7 gigahertz processors nearly two years later. Worse, IBM has also been unable to deliver a G5 that runs cool enough to put in a laptop. Given that laptops now outsell desktops in computer sales, Apple is watching its flagship PowerBook line stagnate and can do precious little about it.
This isn't to say that gigahertz is everything. The G5, though running at lower clock speeds, held its own in many tasks when compared to a higher-clocked PC, and even bested PCs at many tasks (especially involving PhotoShop), and Apple worked hard to make people understand that it was the performance of the entire system that satisfied users, and not just clock speed. Indeed, Apple scored quite a coup when Virginia Tech built what was at the time the world's third fastest supercomputer out of an array of 1100 commercially bought Power Mac G5s (and now using G5 XServes). Other major projects have since been born out of the success of the powerful (and affordable and efficient) XServe, including the "MACH 5" supercomputer built for the US Army out of 1566 XServes and running at 25 teraflops.
Nevertheless, there were clearly obstacles that IBM was unwilling or unable to clear. Worse, with IBM now contracting with Sony, Microsoft, and Nintendo to supply processors for their wildly successful gaming consoles, Apple is no longer a priority in terms of design, supply, and customer satisfaction.
Hence Apple's decision to gradually begin using Intel processors, as announced by Steve Jobs on June 6, 2005, at his keynote address at the World Wide Developer's Conference. Jobs revealed, as long had been suspected, that every release of Mac OS X has been compiled for both the PowerPC processor (used in current Macs) and for the Intel x86 processor (this build code-named "Marklar," after an episode of South Park where everything on the Planet Marklar is referred to as "Marklar.") ... just in case. This makes sense, given the platform independence of the UNIX underpinnings of OS X, and a natural desire to always have a "Plan B" when beholden to another company's whims and proclivities. While the ultimate specifications of the Intel Macs remains speculative (the first systems are a year away, after all) Apple is providing developers with transition kits (including a 3.6 ghz Pentium IV system) that run an x86 version of OS X version 10.4.2 so that they can begin producing software that will run on the new Macs. The strategy at present is to produce "universal binaries" that will run on either PowerPC or Intel-based Macs. Additionally, Apple will be including a translator program, known as Rosetta, that will allow the Intel Mac to run software designed for a PowerPC Mac. The transition will start in 2006 with the consumer level Macs (the Mac Mini and the iBook), eventually moving to the Power Macs and PowerBooks in 2007, followed by the Xserve.
In this respect, I agree with the reasoning of Jon "Hannibal" Stokes in every particular:
So is this the end of Apple? I hardly think so. I spent three hours yesterday reformatting the hard drive and reinstalling Windows Me on a friend's laptop because it was so infected with spyware and viruses that it could no longer even connect to the internet without seizing up. I then had to install a firewall and anti-virus software so that it wouldn't happen again. Conversely, I've never so much as reinstalled OS X on my iBook since I bought it in 2002. Its built-in firewall renders it virtually invisible to even the most tenacious intruders (according to security websites which probe such things), and it is immune to mail worms, spyware, and computer viruses. That's all in the software -- well designed, secure, reliable, and constantly being refined -- and is something that won't change when Macs use Intel processors.
It simply cannot be that a Mac is so connected to the processor that it will be indistinguishable from a PC once both use Intel chips. Macs have undergone significant changes without losing their identity. The move from 68000 series processors to the PowerPC chip in the 1990s also required people to buy new software and new systems, and required emulation to run legacy software, and it was necessary to do so for the company to remain viable (notwithstanding the terrible management in the Scully and Amelio years that almost ran Apple into the ground before Steve Jobs returned in 1997). Windows computers are very focused on legacy support. Apple, conversely, is an early adopter of new technologies. A PC today could run DOS and other software from 10 years ago. A modern G4 or G5 Mac could not boot System 8.6 -- the newest Macs cannot even boot OS 9.
What's in a Mac anyway? The same hard drives as can be found in any computer; the same CD drives; the same USB and Firewire ports; substantially the same RAM; and so on. The main difference has been, of course, the instruction sets for the processors, requiring custom video cards and lacking BIOS, but given the radical changes to Macintosh design over 20 years, a Mac is a computer that runs the Mac OS, is well-designed, and just works.
As for me, I'm hoping to somehow have it in my budget this year to buy a new G5 iMac (and finally retire my old PC running WIndows 98), but in the meantime the law firm I'll be working for only uses Macs and so I'll finally have Mac goodness both at home and at work. So that's all good. And if I can't buy a new Mac until they're using Intel processors, well, I'll be perfectly happy to get me one. Betcha they'll be better than ever.
Comments:For years, of course, Apple's computers were based on the Motorola 68000 series processor, followed by the PowerPC design. Most recently, Apple began buying its coveted G5 PowerPC processors from IBM, but has been stung by IBMs ability to provide these processors in great volume. When Steve Jobs introduced the G5 processor in the first 2 gigahertz Power Macs over a year ago, he proudly stated that the G5 would have a clock speed of 3 gigahertz within twelve months. This never materialized, and Apple's top of the line Power Mac G5 has two 2.7 gigahertz processors nearly two years later. Worse, IBM has also been unable to deliver a G5 that runs cool enough to put in a laptop. Given that laptops now outsell desktops in computer sales, Apple is watching its flagship PowerBook line stagnate and can do precious little about it.
This isn't to say that gigahertz is everything. The G5, though running at lower clock speeds, held its own in many tasks when compared to a higher-clocked PC, and even bested PCs at many tasks (especially involving PhotoShop), and Apple worked hard to make people understand that it was the performance of the entire system that satisfied users, and not just clock speed. Indeed, Apple scored quite a coup when Virginia Tech built what was at the time the world's third fastest supercomputer out of an array of 1100 commercially bought Power Mac G5s (and now using G5 XServes). Other major projects have since been born out of the success of the powerful (and affordable and efficient) XServe, including the "MACH 5" supercomputer built for the US Army out of 1566 XServes and running at 25 teraflops.
Nevertheless, there were clearly obstacles that IBM was unwilling or unable to clear. Worse, with IBM now contracting with Sony, Microsoft, and Nintendo to supply processors for their wildly successful gaming consoles, Apple is no longer a priority in terms of design, supply, and customer satisfaction.
Hence Apple's decision to gradually begin using Intel processors, as announced by Steve Jobs on June 6, 2005, at his keynote address at the World Wide Developer's Conference. Jobs revealed, as long had been suspected, that every release of Mac OS X has been compiled for both the PowerPC processor (used in current Macs) and for the Intel x86 processor (this build code-named "Marklar," after an episode of South Park where everything on the Planet Marklar is referred to as "Marklar.") ... just in case. This makes sense, given the platform independence of the UNIX underpinnings of OS X, and a natural desire to always have a "Plan B" when beholden to another company's whims and proclivities. While the ultimate specifications of the Intel Macs remains speculative (the first systems are a year away, after all) Apple is providing developers with transition kits (including a 3.6 ghz Pentium IV system) that run an x86 version of OS X version 10.4.2 so that they can begin producing software that will run on the new Macs. The strategy at present is to produce "universal binaries" that will run on either PowerPC or Intel-based Macs. Additionally, Apple will be including a translator program, known as Rosetta, that will allow the Intel Mac to run software designed for a PowerPC Mac. The transition will start in 2006 with the consumer level Macs (the Mac Mini and the iBook), eventually moving to the Power Macs and PowerBooks in 2007, followed by the Xserve.
In this respect, I agree with the reasoning of Jon "Hannibal" Stokes in every particular:
I don't really have to go into much detail about the current, pathetic state of Apple's portable line. Even if the Pentium M didn't spank the G4 in performance per watt, I could still sum up in a single phrase the reason that three years later I'm still limping along with the same TiBook: 166MHz frontside bus. This is just unacceptable for a machine with a price tag as high as the 15" PowerBook. In marked contrast to the currently stagnating G4, the Pentium M is fast, and Centrino is a great, full-featured, low-cost mobile platform that just kills anything that Apple could hope to offer based on parts from either IBM or Freescale. Apple needs the Pentium M in its mobile line, and it needs it yesterday.
The second reason why it makes sense to introduce x86 via the portable and low-end Macintosh lines is that neither of those lines have any need for a 64-bit processor. Yonah (see below), which is the dual-core Pentium M derivative that Apple will probably put in its first x86-based PowerBooks and Minis, will not support x86-64. By the time x86-64 has spread widely throughout the Pentium desktop line at the end of 2006, Apple will be ready to introduce 64-bit Pentium-based PowerMacs.
In this respect, Apple's x86 platform shift strategy is deliberately the reverse of Intel's 64-bit platform shift strategy. Intel is introducing 64-bit support into its products from the top down, with the mobile processors not getting 64-bit support until late 2006/early 2007. Apple, for its part, already has just such a 64-bit workstation/32-bit mobile split with the 970/G4 pairing. So Apple can swap the 32-bit G4 for Intel's 32-bit Yonah, and gain an instant performance boost where they need it most without sacrificing a prominent feature like 64-bit support. Later, as Intel moves to 64 bits across its entire desktop line, Apple will upgrade its existing 64-bit PPC parts with higher-performing 64-bit Intel parts. The end result is that as Intel makes the transition to 64 bits, Apple will make the transition to Intel.
So is this the end of Apple? I hardly think so. I spent three hours yesterday reformatting the hard drive and reinstalling Windows Me on a friend's laptop because it was so infected with spyware and viruses that it could no longer even connect to the internet without seizing up. I then had to install a firewall and anti-virus software so that it wouldn't happen again. Conversely, I've never so much as reinstalled OS X on my iBook since I bought it in 2002. Its built-in firewall renders it virtually invisible to even the most tenacious intruders (according to security websites which probe such things), and it is immune to mail worms, spyware, and computer viruses. That's all in the software -- well designed, secure, reliable, and constantly being refined -- and is something that won't change when Macs use Intel processors.
It simply cannot be that a Mac is so connected to the processor that it will be indistinguishable from a PC once both use Intel chips. Macs have undergone significant changes without losing their identity. The move from 68000 series processors to the PowerPC chip in the 1990s also required people to buy new software and new systems, and required emulation to run legacy software, and it was necessary to do so for the company to remain viable (notwithstanding the terrible management in the Scully and Amelio years that almost ran Apple into the ground before Steve Jobs returned in 1997). Windows computers are very focused on legacy support. Apple, conversely, is an early adopter of new technologies. A PC today could run DOS and other software from 10 years ago. A modern G4 or G5 Mac could not boot System 8.6 -- the newest Macs cannot even boot OS 9.
What's in a Mac anyway? The same hard drives as can be found in any computer; the same CD drives; the same USB and Firewire ports; substantially the same RAM; and so on. The main difference has been, of course, the instruction sets for the processors, requiring custom video cards and lacking BIOS, but given the radical changes to Macintosh design over 20 years, a Mac is a computer that runs the Mac OS, is well-designed, and just works.
As for me, I'm hoping to somehow have it in my budget this year to buy a new G5 iMac (and finally retire my old PC running WIndows 98), but in the meantime the law firm I'll be working for only uses Macs and so I'll finally have Mac goodness both at home and at work. So that's all good. And if I can't buy a new Mac until they're using Intel processors, well, I'll be perfectly happy to get me one. Betcha they'll be better than ever.
For reference, Google gave me:
Results 1 - 10 of about 19,300 for "hell freezes over" apple. (0.19 seconds)
Oh, and to get one of those developer computers, you have to be a gold partner or whatever, which means paying about $500US per year, and then another $1000US for the computer itself. It's too bad. I'd love to buy one. I'd even promise to port some open-source stuff to it. :-)
Results 1 - 10 of about 19,300 for "hell freezes over" apple. (0.19 seconds)
Oh, and to get one of those developer computers, you have to be a gold partner or whatever, which means paying about $500US per year, and then another $1000US for the computer itself. It's too bad. I'd love to buy one. I'd even promise to port some open-source stuff to it. :-)
Thanks for providing some hard numbers, Chris. I was expecting some insightful feedback from you and I'm glad you followed up so quickly.
In general, I'm always surprised at how much attention is paid to Apple by the general tech community, much of whom also insist on pointing out Apple's small marketshare in computer sales. There's a belief that Apple is a bit player, but at the same time is responsible for SO much innovation in an otherwise stagnating industry.
I can understand why many people would be engaged in a fair amount of hand-wringing or navel gazing over this, but if at the end the Intel-based Macs are well-engineered, well-designed, and run with the famous reliability we've come to expect, what's the problem?
In general, I'm always surprised at how much attention is paid to Apple by the general tech community, much of whom also insist on pointing out Apple's small marketshare in computer sales. There's a belief that Apple is a bit player, but at the same time is responsible for SO much innovation in an otherwise stagnating industry.
I can understand why many people would be engaged in a fair amount of hand-wringing or navel gazing over this, but if at the end the Intel-based Macs are well-engineered, well-designed, and run with the famous reliability we've come to expect, what's the problem?
