Originally posted by hvfsl
We are getting a bit ahead of ourselves here. We do not even know for sure Apple is going to use the PPC970, they may keep to the Moto G4 until the time comes to release Mac OS X for AMD/Intel. I just think that people are setting themselves up for a big disapointment if the 970 does not come out for the Mac in 2003, or at all.
I agree absolutely that we are getting ahead of ourselves in anticipating the 970 this year. Apple has not committed to the 970 (and likely will not publicly until the machines are ~1 month away). I am certainly not holding back any hardware purchases for the 970s to arrive, and neither should anyone else. Apple may not take the 64-bit plunge until the 970+ (Power5 variant) comes out for all we know. And as for when the earliest 970 will be, I strongly suspect it will be XServe-only, and I highly doubt that we'll see it mid-year. I suspect we'll get at least one full rev of XServe before the 970-based units debut. On Apple's Power* lines? Early 2004 is the easrliest I would hope for such; more likely MWNY-timeframe 2004 (late Summer).
Apple only cares about the short term, not the long term. If they did care about the long term they would have switched to x86 chips a long time ago since Apple has been behind Intel/AMD in terms of raw speed since 1997.
Hmm. And yet, the x86 line is reaching an end, while the PPC line still has massive room for growth. So, a "foreward-looking" company would switch ships to the one that is sinking instead of sticking with the one that is bound for shore? Sounds like you are the one that cares only about the short term.
Switching to x86 will mean that Apple will get 20% of the market very quickly, just look at windows and linux sites were they would love to install Mac OS X on there computers, but don't like the low speed and cost of current Macs.
No, Apple wouldn't get that much marketshare. The chief retaining factor for Windows right now is software, not hardware. Too many people and companies need Windows-based software. Were this not true, Linux would have at least 10% non-server penetration, instead of the 1-2% non-server penetration it has now (overall Linux "market share" is higher, but these are primarily servers, not desktops).
Assuming Apple could, through some odd manipulations, gain a 10% marketshare overnight. The cost to Apple would be its entire hardware division and revenues, and its gain would be a second-string OS business on commodity hardware. Second-string OS's have a historically hard time gaining marketshare against Windows. Any additional revenues would be completely offset by the required marketing and promotional costs. OS development would stagnate (and MS would steal the last couple of original ideas they hadn't already incorporated). OS X would go the way of DR DOS, Geos, BeOS, NeXTStep, and OS/2.
This is not forward thinking. This is suicide. This is stupidity. Apple will not sacrifice its highly-profitable hardware business, a business in which Apple enjoys a considerable advantag over the competition in terms of forefront design and simplicity-of-use integration, for a very short-lived software mini-success. Apple makes good money off software. However, i can guarantee you that the lead it has with OS X and its overall development focus can not offset the war-chest fund difference wherein MS has 10x as much money in cash. MS could easily slash the price of Windows and Office/Windows to $50/per and still make a profit (currently they make 85% profit on average of $200/Office sale and $100/Win sale). Apple can't compete there on price (I suspect they'd start losing money at $100/Marklar), and can't fund feature innovation with their war chest more than MS can.
Head-to-head with MS is not an option.
Most people now agree that Mac OS X is the best OS, but most people still say it runs on the worst hardware.
Not the worst hardware by a long shot. Architecture-wise, PPC is a beautiful architecture, and has a long life left. The x86 crowd is seeing end-of-life within ten years (and substantially decreased vitality until then). The future for Intel/AMD will be Itanium or the like, or a frankendesign based around AMD's Hammer architecture (which compounds the worst problems of x86 in that it keeps backwards compatibility with a poorly-designed instruction set and keeps consistent Intel's longstanding x86 bugs, but which has a higher chance of surviving simply because it doesn't require old software to be tossed out) or Intel's Yamhill architecture.
Should Itanium win out, we have a tremendous opportunity for PPC: old programs have to be replaced anyways, so moving over to a time-tested and well-populated application space on PPC becomes less of a shock to the status quo. Should Hammer or Yamhill win out, the PPC will see relative performance growth greatly exceeding that possible on the x86 compormise designs, and will shortly be back in the performance lead. Should the battle remain undecided for a few years, PPC wins because it will continue to have a much larger application base and a much more stable and known future than either of the "mainstream" contenders.
On the other hand, were Apple to switch to x86 or it's successors, we'd have to pick (today) the winner in an as-yet-unfought fight. Would Apple throw in with Itanium? How about Hammer? Yamhill? We have at least a 66% chance of betting on the wrong horse, and at a tremendous opportunity cost.
Switching to x86 today would be utterly stupid. Period.
