Right, but the keynote is marketing. And they've been saying that for, what, nearly a decade now? [0] And in that time they have chipped away at that segment from different directions, with the iMac Pro and the Mac Studio. The latter of which is a beast already.
What's left? Render farms, scaleable computing.
I guess I think there will be some sort of new machine in that segment. But I figure it's not going to be what people expect from a single, unitary Mac Pro desktop.
Because once you're up at that level of performance requirement, you start to want modular and scaleable things.
What is the pitch for the part of the market that is used to using commodity hardware in scaleable configurations?
I guess. I mean, it's a forgettable stopgap machine (I forgot it ;-) It's not really what people wanted, is it? It's just something they had to put on sale to shut people up. A conventional, Intel-era thing.
They've been saying "something really great is coming", all that. That Mac Pro was not that.
Does Apple still care about the market that wants to plug in internal GPU cards? Does that make much sense even in the context of hardware the speed of the Mac Studio?
What? It's a whole new tower with PCIe slots and everything, and was super well received. Were you still under the impression they were selling the trashcan Mac Pro?
I simply forgot that they made it. I forgot a lot of stuff in 2020 and 2021, like most people.
But by extension I certainly don't remember it being "super well received", or I'd probably have remembered it at all.
Either way: look at it. It's a tower PC. Do you think the current Apple trajectory has any meaningful room for that? What are people going to really be putting in it except disks?
It's an amazing machine.
It's an incredibly expensive machine, but still amazing.
> Do you think the current Apple trajectory has any meaningful room for that?
Are you upset that you think Apple is going to stop making something that you forgot they made, except there is absolutely 0 evidence that they will stop making it?
We know for a fact that they are going to release a new Apple Silicon based Mac Pro.
We knew Mac Pro would be the last thing they move over to Apple Silicon.
Apple even gave a timeline when they first launched M1, and so far it seems to be on track.
Apple has been extremely pragmatic lately, backtracking on objectively bad decisions around everything from keyboards and ports to form factors.
I didn't phrase it that way, but thanks, I guess. Whatever.
> Are you upset that you think Apple is going to stop making something that you forgot they made, except there is absolutely 0 evidence that they will stop making it?
I didn't say they'd stop making it, quite.
So much as that I think the product designation is a phantom, in a way. And the more alternatives they add to the Mac Pro, the less the market needs it.
How many people who bought the forgettable machine put an expansion card in it, do you think? How many of those people thus simply do not need anything more than the Studio?
And also: how many third party manufacturers are going to rush out of the gate to port their drivers to some complex new multiprocessing M1 machine?
I agree they have been pragmatic. But I think they've also sliced and diced this segment to the point where it doesn't make the sense it did.
If all you want is disks then the Mac Studio seems to be exactly what you'd want. I would assume people that want a tower want exactly what you were asking for above, which is GPUs or other cards like audio engineers need.
I didn't see a single negative review of the tower at all, so I'm really not sure where that comes from. And they said during the keynote today a new one is coming.
If their GPU can do workloads replaceable ones can’t then it’s worth it. If you need that amount of GPU, you’re doing high end shit with a high end budget, and can afford to just replace the machine in 2 years (if you even need it - many people stick with their Mac Pro 5-10 years).
Tons of people in pro industries bought it and love it. Pre-COVID it was sold out for almost a year. Every professional music producer I know bought one, and I’m sure they’ll get the M1 Mac Pro.
Funny story. Max Martin partnered with Creative to make a new audio rack to enable the audio for the first Back Street Boys album. That got miniaturized to audio cards for gaming which further shrunk down to a DAC and that lead to the iPod which lead to the iPhone which lead to Apple Silicon which brought us here today. Music producers are huge fucking tech nerds and literally move technology forward. Apple frequently works with producers when developing their highest end products that eventually trickle down to consumer grade.
Looking at the Mac Studio... what would an M1-based, drop-in Mac Pro replacement offer on top? That sounds like a tail-end product.
It's not like 2022 Apple is the kind of business that announces deep partnerships with graphics card makers.
It feels more to me like we're going to see either something that is half-rack-half-Mac, or something with major developments in neural engine hardware, or something.
A new segment -- something that is going to need significant new OS work.
CGI, CAD, rendering, driving huge screens, running live sporting events, scientific workloads, massive simultaneous automated app testing, etc. Just because you can’t imagine why someone would need that much compute doesn’t mean the use cases don’t exist.
I did not say -- at all -- that I can't imagine that kind of power requirement.
Not at all.
What I said if you read the rest of my comments is that I find it difficult to believe there is a specifically single-desktop-machine use case for that kind of power over and above the Mac Studio.
The reason I mention this is that high end (TV and cinema) CGI just is not the domain of single machines anymore; it's the domain of render farms. The people who do that kind of work for TV and cinema now expect to run farms of commodity hardware that can be swapped out and replaced. And that technology is available to everyone in a way that can be constructed more pragmatically.
Scientific workloads, similarly: most of that market is not going to spend a bucketload on a single Mac when they can spread their risk with cluster computing.
App testing: again, an application for clustering, and low cost hardware spreads risk.
So my point, again: given the existence of the iMac Pro line, and the M1 Ultra Mac Studio (with its evident astonishing GPU performance), given that I imagine most Mac Pro users never put an expansion card in their machines (which is the grand theme of Apple -- people don't upgrade or expand, usually), and given that cluster hardware is commonly in use and well-supported, is the niche for single mega-expensive desktops really big enough?
I don't think it is -- you think it is. But I think you can disagree with me without imagining me stupid, as I disagree with you without doing the same.
The real-time rendering demos of the Studio were incredible. If the pro is 4x that it would be insanely cool. If I were still doing music videos I’d kill to have one of those on set. Or what we did for Kanye’s Yeezus tour (live motion tracking of the dancers with some kinects and putting them into models projected on the screen) - could’ve been so much cooler with this much compute in a small box. You’re talking about this as if you have knowledge from working in an industry that you clearly haven’t. $100,000 fragile rack or $8,000 shoe box you can drop and it’ll still work fine.
The visual arts this thing is going to enable us going to be amazing, and we haven’t even seen the Pro yet. One former client has been texting me all day about ideas from 10 years ago that weren’t feasible but now are with this lil thing. I keep telling him to wait for the Pro then we can take some REALLY neat ideas off the shelf.
> I keep telling him to wait for the Pro then we can take some REALLY neat ideas off the shelf.
Aside from the fact that you're making my point for me -- you think the Mac Studio could do the job you imagine a Mac Pro doing -- there is this:
You tell your former customers to wait for an unreleased, as yet unscheduled update to a product to which historically Apple has displayed considerable, time-insensitive indifference, rather than order maybe multiples of the product that has been announced, or try to make it work somehow with kit that exists?
I'll note down your prediction about $8000. That, I guess, would be interesting. But if it's that cheap it's going to be after the chip shortage is over, surely.
