I'm going to risk a few flames here by saying something I've been thinking since last fall. I'd love to see Apple offer as an addition to the current PowerPC line, an OS X based system running on an Athlon or ThunderBird CPU and PC 6 PCI slot motherboard. Personally I could not care less what CPU the MacOS is running on, and this would be a way to play catch-up on the hardware side (4x AGP with fast writes, 200MHz or better FSB, 133MHz memory bus (or higher), 6 PCI slots and a 1GHz CPU). MHz isn't everything I know but all other factors being equal, higher clock speeds boost all application's performance - not just those that support specific instructions. (I'm not talking about cases where the Video card is the bottleneck, but general applications performance.) This is why MHz/MHz a G3 usually runs non-Altivec applications as fast as a G4 (see cpu upgrade applications tests pages for samples). The stronger DP FPU of the G4 is not a real factor in most common applications that the average Mac owner uses. When you say 'super-computer', consumers naturally expect all applications to run faster, not just those that support specific instructions. The main way that is assured is higher clock speeds. But most of you know that already. Problem is Motorola has been stalled (as far as mass production) at 500MHz for about 1/2 a year now and there may be another 3-6 months before significantly faster G4s are shipping in volume.

And the issue is not just CPU speed, but the motherboard, graphics and system bandwidth as well. If you look at the potential performance of even the current G4 CPU, it can't reach the maximum throughput of Altivec's potential in the current motherboard design (not enough bandwidth). And I won't rehash the limited graphics performance issues that have been beaten to death in the past. To really address that means changes in both motherboard design and CPU. This is another area where the Althon, ThunderBird (or PIII) and a modern motherboard for it (4x AGP, fast writes, higher memory bandwidth, write combining in the CPU) would be a big plus. Hopefully Apple knows something we don't in these areas as far as future G4 cpus and motherboard designs in the works, but insider info I have says there will be no write-combining in future PowerPCs that are in the works now.

If I were Apple, I'd have an Athlon, Thunderbird, or other high-performance CPU running OS X in my lab, just to avoid having 'all my Apples in Motorola's basket' so to speak. IBM is apparently restricted in some way regarding selling the G4 based on info I've heard in the past, otherwise we might have had faster G4s already, but that's a moot point. Regardless, the advancements in processors since last fall have been 'constipated' to put it mildly. Yes, I've heard all the rumors on low power G4s, faster versions, dual processor cores (a rumor heard long before the first G4 was even released), etc. etc. but that's nothing but talk so far. Bottom line is it will be months before any of these versions are available, maybe even the first of the year. That's a long time for a consumer market that can buy 1GHz systems at retail today and 900MHz CPU modules for $580 end user price right now on the net. By the end of the year the hardware gap will widen even further. That doesn't matter to many people, but I hate to see Apple lose even one customer, or potential customer. And the whole issue of stagnated CPU advancements have hurt the Mac's image worldwide.

Motorola and IBM are just not as focused on consumer/general purpose high-performance processors nor do they have a huge market to sell them to. They can't afford to spend the billions in R&D that AMD and Intel are doing and will continue to do in the future. And the performance curves in some areas are going to start looking logarithmic pretty soon. I'm not talking about just CPU speed, but graphics performance, memory, storage interfaces and more. Looking long term, I think it's time Apple considered other alternatives to Motorola/IBM, especially since OS X can be run on other processors. Apple has too much to lose to pin their future on one CPU design from a company that can't seem to get off TDC at the moment. What happened to the 1GHz PowerPCs that IBM was to have this year? No matter, as there are still issues of system/graphics performance that need to be addressed, both at the CPU and the motherboard. By offering an OS X Athlon/Thunderbird system, you instantly have access to 4x AGP/fast write (writes directly into the graphics card's ram), write-combining, motherboard designs with higher memory bus performance and more. These motherboards cost $150 or so at quantity 1 retail, so imagine the cost in volume. I know this opens up the potential for another mixing hardware nightmare as is often seen on the PC, but we can't keep using that as an excuse to close the box up, limit performance and choices. And look how Intel has assisted PC manufacturers over the years with subsidized ads (untold millions), hardware design, coding support and more. Apple never had the luxury of a CPU mfr. that poured money and resources to actively help them succeed. Just as Intel did with Linux, I suspect if Apple actively started working with AMD hardware Intel would be knocking down the door with cash and resources to help get in on the act.

Again, I'm saying this could be an addon to the existing line. I am not suggesting Apple mess with the existing models that are selling well and fit many user's needs perfectly. What I am saying is to go after the performance market, which I suspect is a lot larger than Apple realizes. Tim Seufert, a long-time Mac owner and advocate put it very well the other day in the forums. I wish I had saved his exact words. It was basically that he loved the Mac design, ease of use, etc. but also would love to have a Mac that was second to none in performance in all areas. I'd not be writing this if I didn't care about the Mac and before you flame me ask yourself if you spend hours every day and night trying to help other Mac owners, day after day after day. I do, and I'm only writing this because I think it needs to be said for the long term good of the platform and Apple as a company. (And in case anyone is wondering, I'd not trade my G4 for any 1GHz PC. I could have done that months ago (traded for a 1GHz Athlon) but didn't hesitate for a second in turning it down.) While on the subject, I'd not trade my Powerbook for 3 of any PC notebooks ever made.

For these systems, Apple would also save all the costs of designing, developing and testing their own ASIC/motherboard designs, while being able to quickly take advantage of changes in technology over time. And with PC CPU prices dropping dramatically every quarter, their system pricing and features could be updated often. Some bring up the issue of endian differences in PC/Mac (noted in the FAQ here since 1997 and covered again last summer in discussions on TNT cards), but remember Linux was ported to the Mac and both ATI and 3dfx have firmware/drivers for both platforms so the endian issue is one that has been dealth with before. I know Apple is strapped now with OS X (PowerPC) but I still think this is worth considering. (I hear Apple is hiring PC engineers and Darwin for X86 exists in beta form.) The pros far outweight the cons in my opinion. And to those that have been bitten by the 'rev 1' motherboard bug, you could replace it easily (if a bios update did not fix the problem) for less than $200.

There's also another factor to consider - imagine the potential increase in Apple OS users if there was a PC hardware compatible version of the system software for sale. I've read that many years ago, Apple had Mac OS running on a PC when Windows was at version 1.0. It was scrapped I hear due to the thinking that 'hardware is where the money is' (remember when IBM said this and Gates laughed all the way to the bank?). If they had released MacOS for PCs at that time I'll bet Mac OS would be the standard now, as Windows versions up to v3.0x were *horrible* in design, interface, and stability. That was a *huge* opportunity missed that could have changed the course of computing history as we know it. If I were Apple I'd think 'outside the box' and really shock the world while realizing a ten-fold increase in installed hardware that would be able to run the Mac OS. It would also help ensure Apple's future is not at the mercy of future PowerPCs, and just might spur Moto and IBM to do even better. Of course there's also the risk that they would negatively react to this move, but to do nothing and have them realize they've got the Mac market by the throat is not a good thing for Apple's long term success and growth. Things look rosy now perhaps, but Apple cannot rest on its laurels or get complacent. They have to aggressively plan ahead for the long-term, and carefully look at what advancements are being made on the other platforms.

The advantage of Apple selling a system with OS X already installed is that it could be certified for use with OS X (avoiding the support issues of selling the OS software only to a world of mixed PC hardware - and remember MS did well with NT even with its limited list of compatible hardware). But if Apple wanted to avoid selling PC hardware, perhaps a Linux-like release of an 'unsupported' (and low-priced) PC hardware compatible OS X version might be an alternative. Either one of these choices would be a very good thing for them to consider as far as a long-term vision. But, as Dennis Miller says - that's just my opinion, I could be wrong.
-Mike xlr8yourmac.com