Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Nobody has a poster of a base 911 on the wall. These halo products need to exist,

Sure, just the small problem that no MPro ever was a halo product.
The 2013,19 and M2-Ultra ones were the exact opposite of a halo product.
They were seen as pointless waste of money by anybody who didn't have a specific use case for them. Regardless of Mac owner or not.
 
  • Haha
Reactions: Flint Ironstag
As for Apple being fragile, well, all the players in the highest-performance space at all levels right now are dependent on TSMC in Taiwan.

For perspective; I think Apple is fragile insofar as I thnk their inclination is to double down on a trend, just as it becomes over-cooked, and culture goes in the opposite direction. The Macbook Neo, it's cheap, and there's a worldwide recession. I would be unsurprised if Apple took the success of that product to "improve it' by making it more expensive, putting a notch into it, shuffling it higher into the pricerange, when cheap was literally its only selling point.

To look at the car world by way of metaphor; Ferrari have been going more and more digital with cars, more and more electronic, more and more performance, unheralded performance...

...and what are the customers for supercars and hypercars demanding from makers?


Customers want slower cars, with manual transmission, no hybridisation. Electric supercars are being cancelled, hyper-analogue restomods are the hottest thing in the automotive world. Customers want a visceral experience; a steamtrain they can use on the road.

And Ive now designs a car for Ferrari that's been panned by everyone who already owns a Ferrari, will probably sell some to vulgar new money people, but it'll never become a collector's item, so the financing packages which underpin a lot of supercar sales will collapse for it.

Like Ferrari, Apple is fragile to its own success (and hey, there's Eddie Cue at both companies).

AI is the feature Apple's performance is so good for? But AI is the feature most people HATE in their products, so who is all this performance serving? #NoAI authenticity is the emerging paradigm. In day to day usage, the modern Mac is less capable than it used to be, because modern software has gone so far down the "simple is easier" fallacy, that it's past the inflection point of "less is less".

I suspect KDE, for example, can get good at being user-friendly and covering more people's use cases on existing "obsolete" hardware faster than Apple can become anti-fragile to circumstances. If the EU for example, which is a larger population than the USA, uses competition law to fundamentally break the trying points of Apple's ecosystem (given the political animosity, and Europe's sovereignty push for technology), I think that's existential for Apple.
 
Nobody running towers of this caliber cares about bang per watt. We can afford the electricity. And our power crushes AS. Come to grips with that.
...which is one good reason why Apple is wise to focus on mobile, laptop, all-in-one and small-form-factor systems - which is what the vast majority of people are buying anyway - where Apple Silicon has a clear advantage.
 
...but the M3 Ultra Mac Studio is also perfectly suited to local LLM work, and costs $2000 less (which also makes it viable for clustering). All the Mac Pro offered over the Studio was PCIe slots, which are irrelevant to Unified RAM-based LLM use.


Apple's stated reason for the current shortage & (probably temporary) discontinuation of some higher-RAM Studio was unexpected demand due to LLM use (probably due to M4 and M3 Ultra Studio stocks drying up before the M5 models were due to launch). That shortage is a problem, but a new M3 Ultra Mac Pro would only have made that problem worse.
The difference using a MacPro would be that you don't need to link multiple MacStudio.
You loose performance.
 
AI is the feature Apple's performance is so good for?

The proposal that started this thread was that Apple should have made new Mac Pro with lots of plug-in GPUs for running local AI. The counter-argument was that the Mini and Studio are already selling to people who want to do local AI (to the point that they sold out) without Apple really needing to "double down" on anything (apart from maybe getting the M5 Mini and Studio out the door a bit smartly). The Mini and Studio have much broader markets than just AI and will still be in demand if the bubble burts. Even the Studio Ultra is a "simple" doubling-up of Max dies and is a much lesser gamble than the all-new die that would be needed for a credible PCIe tower.

Even with AI in general - everybody has been screaming at them to up their in-house AI game for a couple of years, but they've instead licensed third party models for their Siri update. Given that, as we speak, markets are falling partly because people are questioning the value of AI stocks, Apples reticence to go big on home-grown AI could turn out to have been a company-saving throw. Meanwhile, they've have to mention AI every five minutes to stop the shareholders panicing.

They'd be mad not to follow through on the success of the Neo and dumb to make it significantly more expensive (although computer prices are being pushed up across the board) - but there's no sign yet that they will be so stupid. Actually, a cheap Neo gives them some scope to push the Air a bit more upmarket.

Note that the Neo was not random accident - it's a clear reaction to the success of the "new old" M1 Airs that Apple were selling via Walmart up to the Neo launch.
 
  • Like
Reactions: Omega Mac
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.