Was it Apple's choice or IBM's?

The G5 is a great chip

The G5 is a great chip TODAY, but if a 3 GHz version is no where in sight, you have to ask yourself: where will Intel be in a year? 4.5 GHz, at least, I think. Though I never liked the X86 CISC architecture, at least Intel comprehends that chips are its mission in life. MOT never grasped this. They kept their heads filled with little cellphone thoughts, always missing the big picture. And IBM? hesitate and all is lost. Without Apple's high-margin G5's defraying the cost of running expensive chip fab plants, I don't think any number of sales of low margin XBox/PS3/Nintendo chips will ever make up the gap. I sold my IBM june 6, because the other parts of the company just don't justify my continued interest.

According to the Apple

According to the Apple document at

http://developer.apple.com/hardware/ve/summary.html

"Executive Summary

What is AltiVec?

AltiVec is a Single Instruction Multiple Data (SIMD) machine and an extension to the current PowerPC instruction set, designed to improve the performance of any application that can exploit data parallelism. Examples are 3-D graphics (Electric Image, games), image processing (Quartz, Photoshop filters), video processing (MPEG, MPEG2, MPEG4), and theater-quality audio (Dolby AC-3, DTS, mp3), and high performance scientific calculations. It is present in all G4 and G5 machines that Apple sells.

Why do we need AltiVec?

AltiVec offers greater flexibility and opportunities for better performance in video, audio and communications tasks which are increasingly important for applications. AltiVec provides a cornerstone for robust and powerful multimedia capabilities that significantly extend the PowerPC instruction set.

How do AltiVec Features Compare with MMX, SSE & SSE2? ..."

Altivec can perform 8 FLOPS per cycle and SSE can perform only 2 FLOPS per cycle. This is like saying that Altivec is an 8 cylinder engine and SSE is a 2 cylinder engine. So Altivec can perform 4 times as much work per cycle as SSE.
That is one huge work capacity difference, at the vector CPU level.

The obvious implication is that SSE would need a clock 4 times the frequency of Altivec to perform the same number of FLOPS per second. The clock speed of the Pentium chip is not 4 times the clock speed of the PowerPC chip so the Altivec is faster at the level of vector CPU operations per second.

Of course application speed depends on many other factors such as motherboard architecture, I/O architecture, compilers, operating system software and application software.
The benchmarks at
http://developer.apple.com/hardware/ve/summary.html
demonstrate that the PowerPC is faster than the Pentium for those applications, with those hardware, compiler and applications configurations. Benchmarks are soon out of date. It will be interesting to see a new set of benchmarks for Apple applications on the recently (2005/06/06) announced Mac Intel Pentium systems.

Has anyone actually noticed about this "video, audio and communications tasks which are increasingly important for applications. AltiVec provides a cornerstone for robust and powerful multimedia capabilities that significantly extend the PowerPC instruction set." that we are talking about Apples most important market niche (multimedia, entertainment industry, education, graphical and scientific applications ...).

It will be very interesting to see where they are going to find the extra FLOPS/second for these applications that are seriously demanding in the vector floating point domain.

The timing is right. More

The timing is right.

More than any point in its corporate history, Apple can make a safe and quick switch to Intel. All the technologies are in place.

But for Apple to remain competitive, they'll need a lower cost of components, a larger supply of processors and an easier roadmap for software developers. Intel provides all of these, much more that any other processor platform.

Apple's decision (in the 80's/90's) not to port System 7 to the Intel processor has been acknowledged to be one of the biggest blunders in business history. But their subsequent decision not to develop an Apple PC may have been just as big. While nobody could've beaten Microsoft as the dominant platform for applications, Apple arguably had the chance to dominate both OS and PC hardware. Dell wouldn't have become dominant in hardware.

Now that Intel has been adopted, all computer users may finally have the opportunity to load Windows on a Mac (dual boot with Windows) and enjoy an extremely better looking (and relatively trouble-free) PC that anything that Dell, HP, IBM could ever design.

Tiger, OSX, will probably run faster than Windows on the same hardware configuration. Its interesting to note that Apple nor Intel deliberately did not mention which processor chip will make it into the Mac. Maybe Intel is considering a customized P4 that will not have legacy circuitry that's critically needed to run Windows. OSX certainly doesn't need to support older Windows legacy code. That would free Intel to manufacture lower costs and lower consumption processors and enable Apple to deliver an even faster OSX on Intel than Windows on Intel.

Compared to the other major transitions that Apple had to face during the 68K-PowerPC and OS9-OSX, this time around, Apple (as well as the software developers) has the advantage of an actual Apple/Intel box that already runs. It appears that this transition will not pose the same difficulties as the former transitions. Apple is already running OSX on Intel. Previously, developers had to partner with Apple during the transitions. In the process, the developers became the test-bed for Apple.

Although many will say that the move to Intel is another blunder, this may turn out to be Apple's biggest opportunity to bring finally get back into the 10-20% market share. Apple might have it all - beautiful an unequalled hardware design, low component costs, equal processor speed, software and drivers written for Windows that could easily be ported for OSX on Intel, the ability to dual-boot and most of all, the most advanced and elegant operating system - OSX.

I've been reading comments

I've been reading comments on various boards and it's amazing how flustered folks are getting over this announcement. Yes, it's going to painful for some developers I'm sure, but in the long run Apple will end up with a much better product line. The G5's run too hot and the Apple laptops, while very slick, are woefully underpowered due to IBM's inability to deliver a mobile G5. Adobe sweating bullets? I don't think so, they've always developed for both platforms, so they certainly have the resources and knowledge to quickly port PowerPC apps to x86. As far as AMD, they do have some products superior to Intel, but Intel beats AMD hands down when it comes to mobile. If you look at Intel's complete product line & roadmap on dual/multi core (including dual core mobile, to be introduced in 2006) and it's ability to supply a complete chipset solution, I think it's a better bet for Apple. (IMO, Intel engineering is still superior to AMD) Most people don't care what CPU is inside as long as they get good price/peformance. The key thing Apple has and why they like Macs so much is a superior OS that will keep folks buying Macs instead of Windows boxes. Yes, x86 runs Linux and various flavors of Unix as an alternative, but Linux is still for gearheads (most people have no idea what a command line is). The regular folks will now have a real alternative to Windows. Sales wise it will be a little painful for Apple over the next year or so, until the product line gets moved over. I'm thrilled to see this new architecture for Apple!

Intel is NOT Windows Some of

Intel is NOT Windows

Some of your comments are pretty funny - it seems you can't see the forest because all the trees are blocking your view. Most Apple users are only marginally interested in the inner workings of their computer - the important thing is that Macs just work.

I'm a fairly recent convert to Macs and OSX and have many years of experience building, configuring, and using Intel-based PCs with literally all of the major Windows versions, including server versions and a couple flavors of Linux. I've built and upgraded many dozens of Intel/Windows machines over the past 13 years, as well as made repairs for my friends and family.

Of all the problems I've encountered in these 13 years, I can only attribute 2 or 3 to problems with Intel processors, and at least one of those was my fault. The vast majority of the problems are directly related to Microsoft's shortcomings in the design of Windows itself, especially the overly complicated and infamous "Registry" and their convoluted handling of hardware drivers.

When a person buys a computer from Apple, they get a complete package that's reliable, easy to use, and easy to look at. I watched Steve's keynote and was surprised at the work they had done to make this transition easy for everyone.

Talk about innovation - universal binaries! One compiled version of your software runs on either system. If that annoucement had come out of Redmond I would be highly skeptical -- hearing it from Cupertino I know I can bank on it.

That's the reason people do and will continue to buy computers from Apple, regardless of whether it has Intel or IBM inside. People know that they can sit down at their machine and work or play to their hearts content.

It just works.

It's not about the Chips:

It's not about the Chips: Having used just about every system/chip combination you can imaging in my 30+ years of systems engineering/administration, its not the the chips, its the engineering. Just because Apple decides to use Intel chips instead of IBM does not mean that suddenly they are going to start producing cheap junk. Using history as our guide, examine the Apple legacy and you will see that each time they have made a large shift the quality of the engineering remained the same or improved significantly. That is the reason there has been a Mac in my house since 1985 (128K, 512E, SE, IIci, PowerBook 100, PPC601, G3, iMac G4, and now G5). No I am not a Mac bigot and have a Dell P4 server, Dell C610 laptop and Sun Blade 150 in the house as well. But of all the systems I have touched the Mac's have always just worked and I have never had one fail on me yet. (no I don't get a new system every year, I run my machines at least 3-5 years and then pass them down to my nephews/nieces). While it is true that chip design will force changes in system design, it does not mean those changes have to compromise performance or quality. I have no doubt my G5 will serve me well until I am ready to move on to next generation and my software will continue to function.

Headache or

Headache or Opportunity?

I've been pushing Macs to their limits [graphics and publications] for 20 years. Intel-based [Windows] PCs have never been able to cut the mustard for that. I get a migraine just thinking about it.

When I first heard about Apple's rumored, now announced, move to Intel chips, I croaked. To state the obvious, this is going to affect Apple's traditional vertical markets, no matter how it's handled — and this is nothing like going from Motorola to IBM.

Then I read the text of the actual announcement. What's between the lines may be more important than what was actually said. Some of the things that occurred to me include:

• It sounds to me like IBM is refocusing their resources.

• IBM's recently announced sale of its PC group to a Chinese company has to have played a part in all of this. Ditto for the ramp-up for the consumer use of G5 chips, which can't be cheap or instant.

• Assuming IBM is still willing and able to provide chips to Apple, at whatever negotiated price, where would they be made? How reliable would they be? Would they show up on time? What about technology transfers, cloned chips and blackmarket/graymarket chips? [I would love to have been a fly on the wall for that conversation.]

• How much attention are they willing to give to Apple's G5 chips — and to their upgrade path?

• Or is this something more basic — the amount of heat generated by the G5 chips? They're probably not going into super-thin laptops anytime soon, which doesn't make a lot of Mac people very happy. Maybe IBM couldn't figure out how to make them run cooler or how to make them smaller? And, we still haven't seen a 3.x GHz G5 in the big box, with or without a radiator.

• Maybe Apple sees an opportunity in the new stacked processor chips [in pairs — or more]?

• Have they talked Intel into adding or modifying capabilities to meet Apple's particular requirements? If so, who benefits? Just Apple, or everyone else as well?

• Who's designing the rest of the chips in the set? If Intel does any of it, will they make it exclusive to Apple?

• Could it be that Jobs sees an opportunity to get Macs into more PC shops, with the goal of eventually replacing Wintel boxes with Mactels? Neat trick if he can do it. [Gates must have thought of this by now.]

• Of most importance to most users is how this affects usability and functionality.

• What happens to the operating system, the application software and the SPEED at which real work gets done?

• Can a great operating system get past the x86 limitations? What about bugs in all that rewritten software?

• Are we talking OS XI with a transition similar to that of OS 9 to OS X? If so, what about the legacy G3/G4/G5 systems and their OS X software? Does this define the lifespan of Tiger? Does this mean Tiger is the last OS for the IBM chips?

• Keeping up is one thing; obsolescence is something else. Worse, there's bound to be a problem explaining to the relatively unsophisticated guy with the checkbook why a Mac is better than any other Intel-based PC. It's hard enough now.

• HP and other peripherals manufacturers have to be trying to figure out how this is going to affect all of their printers, scanners, etc., including direct-to-plate prepress.

• Will all of their drivers have to be rewritten or will all of that equipment suddenly become obsolete? Obsolescence isn't going to win friends and influence customers.

If I were at Silicon Graphics, I'd be thinking about how to exploit this situation. They lost a lot of their "low end" market when the PowerPCs came along. Maybe this is their opportunity to take back what was lost, and then some, especially if they can do some wheeling and dealing with the major software developers.

If I were Adobe, I'd be sweating bullets.

In actuality, the most I have to worry about is which hardware and software my inhouse marketing, advertising and art departments will use.