To me the significant changes are:
- 4 Hyperthreaded cores available on laptops, mini's and iMacs means that the multi-threaded performance of these consumer computers now exceeds the capability of all but the most specialized software
- Thunderbolt delivers high-speed mass storage expansion to a computer of any form factor... a feature that was previously only available in the Mac Pro.
Both of these developments in the latest updates to Apples consumer computers are relegating the need for a Mac Pro to fringe workloads.
This is how I see it as well.
Again, I'm not seeing this change in computing. The consumer machines and the pro machines have always had this overlap for the last two decades. Why are we freaking out about it now?
The Mac Pro is designed for workloads where improvement can always be had. By definition, quad core consumer machines don't change this.
It will come down to cost/benefit for each user.
Consider SP users that only need 4 cores....
In the past, even when Apple's other products closed the gap in terms of core counts (when consumer parts hit 4 cores on a single die), they still fell far short in areas such as graphics and storage (i.e. consumer ports such as USB and FW couldn't cut it, nor could embedded GPU's).
But with the introduction of TB, this has closed this gap as well, particularly with storage (I do still see limits with graphics in TB's current revision, but that can be dealt with, such as a single graphics slot in an iMac).
Now when users see this and the cost difference between these systems, they may seriously reconsider their position on a MP, particularly in our current economy.
So the days of the MP being the "right" machine for users that only need 4 cores is likely to decline as a result. I'm not saying I expect the sales decline to happen all at once, but over say a couple of releases. At which point, Apple will have to make some decisions.
They've a few options, such as either EOL the MP entirely in favor of say an iMac + TB + user upgradeable GPU, or change the MP as we know it into a SP only incarnation of some sort (no more DP models).
You may not see this as a drastic change, but users that actually need more than 8 cores would (i.e. animators and any scientists that have heavily threaded OS X applications).
The cost of building the Mac Pros doesn't matter because they make a profit on them. That makes about as much sense as saying BMW is about to leave the car market because BMW cars are very expensive to make.
Of course it matters, as that's what they base their MSRP's on. So if the total cost per unit suddenly jumps a $1k, then that could exceed the point where users are either willing or able to buy. Profit could disappear, as it's based on a minimum number of units sold (i.e. expect n units sold, and get say 70% of that figure could not just reduce the margin, but actually end up a loss).
More importantly IMO however, is declining sales that would result from any notable increase in MSRP, even if it's technically still making a profit (doesn't need to be as drastic as $1k, but it will still end up in a reduced margin). When this happens enough times over future models, that profit will eventually reduce to the point Apple will decide "enough".
I'm not sure you've fully accounted for this issue on future MP sales.
2 years ago, there was a huge benefit moving from a MacBook Pro to a Mac Pro for Photoshop. Not any more.
This is how I see it as well. Not to say they're equal, but the gap has been closed quite a bit.
Then consider the remaining issues could be addressed over the next couple of years, it's possible the MP could be EOL'ed.
I don't know if 6, 8, or 12 cores are fringe cases. You're talking about science, video editing, pro audio, and development, which already were the primary market for pro users.
Most of the threaded professional software out there is from the creative area (video/animation/audio).
In terms of science, what I've seen is primarily limited to chemistry and bioinformatics. A few astronomers use OS X on their laptops (also dual boot Windows), but the grunt work is done on Linux boxes. Engineering has it's subsections of suites as well, but all in all, it's not much when I think of the entire market.
As per development, since OS X isn't the biggest part of the OS market, I don't see it as accounting for a drastic number of MP sales either.
Now when I think of the entire professional market, the number of threaded applications is actually small. So "fringe" makes sense as a general rule.
As per the MP specifically, that will depend on exactly what application/s within a suite every user uses, and how often. We don't have detailed information, but Photoshop users are a big chunk of MP buyers from what I've gathered from MR members, which tends to make us think that statistically speaking, most can't utilize all of their cores most of the time.
The issue could be with OSX, although it might just as well be that apps are primarily developed for Windows and ported to OSX as an afterthought.
Unfortunately, this is the likely cause.
Windows is a much larger market, so for developers that create Windows software, that version gets priority. Due to financial reasons, it tends to be ported to other OS's rather than developed specifically for it (optimized).
Like I said, the focus of Apple has shifted to the iPhone, iPad, etc (the mobility market). This is where Apple sees the future.
I see Apple maybe giving the Mac Pro one or two more updates and then dropping the line, just like they did with the XServe. Even Steve said they weren't selling well, and it was just a small fractional portion of all Apple sales. The Mac Pro has taken a backstage position in the mobile-based lineup of iPhones, iPads and notebooks that Apple is pushing out like candy for all the consumers to gobble up.
The consumer side generates a lot more money, so it's not unreasonable for it to get most of the focus (meaning internally = development funding).
The MP doesn't get much as-is, as the ODM does almost all of the work (little bit of software development since the case hasn't changed in how long?).
But with new tech up in the next couple of years and price increases, the MP could go into a downward spiral (reduced sales = increase MSRP = further reduction of sales ....) until it's gone as we know it or just gone.
