The tone of your other posts indicates that such issues are acceptable in the enterprise market.
Before I got into software programming, I worked in the Apple enterprise market for years. As bad as Apple was, our Microsoft hardware/software was always worse. Bugs are part of working in enterprise. If you're looking for a "bulletproof" system, you're probably looking at servers that are multiples of $10k, not a Windows or Mac server.
It's just the way things are. Would I like everything to be bug free? Yes. But for every annoyance I can think of with OS X/OS X Server, I can think of one on Windows.
No, they weren't the primary reason for the change, but they still had to be taken into consideration (all their computer products had to be in order to be sure the change would be viable from a financial as well as technical POV).
I'm sure they were taken into consideration, but the G5 wasn't terribly hurting. Not nearly as much as the Powerbook G4's. As I said, they were the last machines to switch over. It definitely wasn't a priority. There was even one more G5 revision
after the Intel switch was announced.
No they don't do it themselves, as they chose to outsource it. But it's not free. They pay to have it done.
If they do pay, it's not much. My impression is that Intel does it as a service to Apple.
The change-over reduced their R&D, as they could use ODM's for the MP (Intel for boards for 2006 - 2008, and Foxconn for assembly, then later moved to Foxconn for it all). Companies like Foxconn have more experience with Intel parts (or AMD) vs. PPC, which means they're better equipped from a knowledge perspective to get the job done on time and within budget.
Uhhhhh... Really? I don't think this argument makes any sense. Foxconn doesn't give a crap about the processor design. Besides, all the manufacturing companies are building PPC's these days.
I mean, it's not like they see a PPC board and processor, and their heads all explode while they try to figure out how to put the parts together. Apple used these production lines for years in building PPC machines.
What you need to realize, is that even with Intel's reference designs, there's still R&D to be done. Additional components for features have to be taken into the design (i.e. FireWire chip, and PWM controllers not specified in the Intel designs), as well as board layout (PCB work), parts selection (BOM = Bill of Materials, as there's more than just Intel parts on the board), and validation/verification testing to confirm it all works. And that's just the hardware....
Which is all done at Intel.
I'm honestly not sure where you are going with this. Yes, designing the Mac Pro possibly costs money. But the line is profitable. Soooooo, as long as the line is profitable, why does the design costs for the Mac Pro matter?
Different designs and different parts were used (not just the CPU changed; disk controllers, Ethernet, audio,... were all by different vendors = different design = drastically different boards). The PPC designs would have cost them more in R&D as well, as there wasn't that many OEM/ODM's that had the experience to do it all that well (make time and budget requirements).
Yes, Apple did have to design the PPC boards in house, but they certainly weren't doing it by themselves...
Honestly, you keep talking about R&D, but this sort of R&D is usually the cheapest part of getting a product to market.
Which isn't even close to what I said.
Even if there was no price break on CPU's, there would still have been a profit. But the discount that existed at that time increased that profit by whatever those savings per system were.
I don't think you get what I'm saying...
Apple likely wanted to make waves in the market when they entered. They knew that during the transition people were at risk of buying PC's instead of moving over the Intel Macs, especially Pros. They likely got a sweetheart deal from Intel to massively undercut their competitors during the transition to keep people on the platform.
Now that the transition is over, prices have returned to normal.
It's not so much preferential treatment, as it is to do with quantities purchased (standard practice when buying components - you get discounts with increased quantity up to a certain point).
Preferential treatment is getting parts early.
Unless you're Apple. Apple is very widely known to demand discounts on parts.
PPC was different technology. Moving to a different CPU supplier changed the entire circuit design from the ground up, which has a direct effect on costs.
So they are not the same. That shift created a point in which the old metrics no longer applied.
Huh? Not that much of an effect on costs. It's not like Apple was building the processors. Today's Nehelem and the G5 are actually EXTREMELY close in design.
I think you're placing way too much importance on this. Yes, the CPU supplier changed. But what really changed in the circuit design? The CPU and the chipset. While a change, it's not like they had to build a new machine for scratch. The USB controllers, graphics card, firewire controllers, display, disk drives, optical drives, case, wireless card... They were all the same components.
Again, hardware R&D is usually the smallest cost. The XServe was cut because it was too expensive to keep a production line around, not because it was too expensive for R&D
If it's not making a profit, they don't care. It really is as simple as that.
And the Mac Pro is very profitable. So why are we having this discussion?
And this is the crux.
You're presuming that the enterprise market has a significant market in Apple's economy. But there's no indication that's actually the case. When you look at their published earnings, the profits are primarily generated by the consumer products, not the enterprise lines (XServe, MP, and all their professional software combined - FCP, FCS, OS X Server, ....).
Yeah, I think you're not looking at this right... Think about this one. If you're an enterprise, why are you buying Mac servers? Do you just buy a server and put it in a room somewhere? Or are you going to be buying consumer Macintoshes for people to use with that server?
Mac OS X Server is a driver for enterprise buying consumer machines. Servers aren't independent machines. They're a server, for existing clients, by definition. You don't buy a server without clients, and if you're an enterprise, you certainly don't buy clients without a server.
In the last earnings report call, Apple reported massive growth in enterprise sales. You can bet all those companies will want a server for their consumer machines they just bought.
Any company in business is out there to make a profit. It's their whole reason for existence, despite any rhetoric.
As per institutional customers ceasing buying the Macbooks, iPads, ..., they'll continue to buy, as those are the consumer products their staff want.
Again, if there is no server, they will stop buying. Enterprise does not buy devices they cannot manage. You're looking at this all wrong. Server and consumer products are not two different categories. In enterprise, server products are a
requirement for buying consumer machines. It's like building a house with no foundation. If OS X Server was discontinued, I know several extremely large Apple clients who would stop buy Apple consumer hardware.
Of Apple products, this makes sense, as the XServes ran hot.
They ran normal for a 1U. The problem is that you can't put a good OpenCL card in a 1U server. And most clusters are moving to using GPUs to ramp up their computing power.
Yet not too long ago, there were those that said the XServe wasn't going anywhere either. Nor was there a peep from Apple or a leak from an employee that surfaced. Yet not too long ago, the announcement was published that it's axed.
Who said that? The XServe wasn't the first server hardware to go, XRaid was, and it pretty clearly cast a shadow on XServe.
I think it was a dumb move, but it's not at all on the same level as Apple cutting their server line entirely.
So it's not uncommon for things like this to stay underground until the last instant as to not affect sales of existing stocks. It's also kept to a precious few people as well, and as secretive as Apple is, those that know would have enough sense not to leak that information for fear of losing their jobs (i.e. wouldn't even tell their spouses, parents, siblings,...).
Despite what you may think, Apple does share information on upcoming machines with enterprise. One shop I worked for knew about the iMac before it launched, and we got numerous dates for the Intel Mac launches.
They won't bring you the machine to look at, but they do tell you general information.
(For example: I knew roughly about the original Macbook before it launched. The shop I worked at actually literally got the first ones off the line.)
But those resources have been reduced from what I've seen, as they've been pulled to address iOS and other consumer based projects.
Again, as I said, Apple has the same number of Mac OS X engineers that they did before. Nothing has been reduced. They're building/buying additional office space for the iOS engineers.
Again, not what I said.
There's a point with price increases that they'll no longer produce a profit (ye olde Supply and Demand economics). And there're indications surfacing that the profits are reducing, and will continue to do so.
Did you miss the mention of 2013 being the point the MP will be at a crisis point due to Intel's moves combined with Apple's pricing?
I've not posed that the MP will be cut tomorrow. But it's not unrealistic for this to happen in the next few years (no redesign to follow the 2013 MP's), unless they make some serious changes to what the MP is in terms of design (i.e. cease using Xeons) or their pricing model.
The Mac Pro would have to take a simply huge cut in sales for Apple to discontinue it. Simply massive.
Like I said, none of what you're saying fits at all with what's going on. Apple is working on a long term, multi year rebuild of FCP, optimizing it for high end hardware, and you're saying Apple is going to cut the Mac Pro in 5 or 6 years? Doesn't make any sense at all.
Apple sells a lot of Mac Pros. A lot. If Intel's prices go up, workstation prices are going to go up across the board, and Apple will continue to sell Mac Pros to workstation users. What you seem to be worried about is people who don't need workstations not buying Mac Pros, and there are plenty of workstation users who will continue paying workstation prices, whatever those prices may be, for Mac Pro hardware.
The workstation user market is an entirely different market than say, enterprise. Hardware costs are honestly the least of our concerns.
Off the top of my head, I can at least go back to the Power Mac 6500 which had a bad ATA controller, among other things.