You say "Making them upgradeable will likely introduce design constrains that will considerably increase the R&D costs and make the already niche product even more expensive" and I fully agree.
The moment you say power efficiency proves you wrong. For those people who need performance, they really dont care about the power consumption and it's a fact that powerful chips require more power and heat. You are only justifying mediocre specs and performance of Apple Silicon after all.
And yet, take a look at Mac Pro 2013 and 2023. I would be surprised if their superb data lead them to create mediocre products. Same for Apple Vision and iPhone Air.
Not only that, Apple seems to favor on-device AI, so more powerful devices are even *more* important to Apple's AI strategy than most companies
Yes, Apple's recent history with Vision Pro, decades of Mac Pro flops, the iPhone Air, (should I go on?) show that Apple's superb data really makes better decisions than we do
I would disagree, in that although many "people who need performance, they really dont care about the power consumption," there are many forward-thinking people who need performance and do care about the power consumption. And even among the many that don't care about the power consumption there are a lot [perhaps most] that hate the heat production and the noise.
Then I would highly disagree. Since Apple Silicon chip cant keep up with the max performance, what's the point of keeping the power efficiency? Mac has no options for high-end and workstation grade performance and if you think Mac Studio can replace those, you must be dreaming.
IMO the question is how Apple perceives the Mac product line. Does Apple see a place for MP? I could easily see Infiniband as part and parcel of some new MP beast, especially as Infiniband is low latency and at 400 Gb/s now allegedly going to >1Tb/s by 2028. E.g. a cheese grater box with Infiniband bus and multiple connection points to link in all kinds of things.
Sorry but anyone who states "...lead them to create mediocre products. Same for Apple Vision and iPhone Air" IMO seriously lacks vision. Pun intentional. Neither product is mediocre, with the AVP in particular being the antithesis of mediocre. The AVP is a superb new tech demo, and selling >100k units of version 1 at $3,500 each is even decent from a sales standpoint.
AVP is a great example of being failure with a wrong vision. Even Apple has no idea what they are doing with AVP since the version 1 is considered as a failure due to low interest and sales. If you think 100K is a lot, it's not. Developers aren't intersted in AVP and AR/VR market itself is totally niche as nobody ever succeeds even for Meta.
Frankly, I wish I knew more about where personal-computer-level computing hardware was going. What will future high-end and workstation grade performance be tasked to accomplish? Given the very rapidly evolving AI, it could be darn near anything. E.g. M3 Studio Ultra with 500 GB RAM is apparently very cost-efficient for certain LLMs. And that is just a Studio and just M3.
No, Apple didn’t scrap it, there was never an Extreme. People who didn’t understand that Apple is not Intel or AMD (does not have to show a steep performance curve from the least to most powerful) thought that there would be something to set a top in product apart, performance-wise. What Apple did instead is bring incredible performance to the lowest and most efficient products such that single threaded performance is roughly the same across the board. That’s something AMD and Intel CAN do, but will never do because their business model demands that they ensure that the lowest performance solutions be FARRRR away from their highest ones.
I agree 100%. If they would market their non bloated "enterprise" edition that has zero telemetry and other crap to regular consumers, I wold probably be leaning that way. Under current circumstances, no way.
If Apple skipped the M4 Ultra generation, then it’s very unlikely they ever saw the Extreme as possibly giving them a return on investment.
Ultra will suffer same scalability problem with out ability expand and provide slots to upgrade. Apple probably will go in the direction of providing ability to cluster multiple Mac studios. One ultra Mac Pro is good but give ability to combine multiple studios, it can be lot more scalable.
We disagree. IMO the AVP is not "a failure due to low interest and sales." IMO the AVP is a superb tech demo and v1 tech product. Repeat: superb. And yes, I think 100,000 x $3,500 is a lot of sales. At least 99% (and probably 100%) of the tech firms in the world will be happy selling 100,000 x $3,500.
Correct, I "dont understand the situation of AR/VR markets," except to understand that the AR/VR space is growing rapidly. Rapidly enough that investing corporate effort in the space is a smart thing for Apple to do.
I have a feeling that the Apple Silicon Mac Pro only got created because Apple promised to do so when they announced Apple Silicon and it would have been been embarrassing if they didn't manage to do that.
Good question, and given that the trash can mac withered on the vine for years and years, how the current mac pro offers little upsides over the mac the mac studio and the fastest processor you can get on the MP is a M2 Ultra, the writing is on the wall - they don't see a place for the mac pro
It will be a worthy goal for businesses if there's profit to be made. The fact that the power requirements of the last few nvidia GPUs show a large increase, is evident that they like most of component makers are more interested in making the fastest products.
It’s clearly only a stopgap/last gasp. The Mac Pro housing was ludicrous enough for the 2019 Intel Mac - keeping it for the 2023 which doesn’t need to run Intel/AMD space heaters is obviously a compromise.
Then it only proves you that know nothing about the AR/VR market. AVP is nothing new but being luxury and being superb has nothing to do with it. Demonstrated? I told you, AVP is nothing new and already demonstrated by other companies. AR/VR market is good ONLY in B2B market, not B2C market. Literally, there are no companies ever succeeded in B2C market including Meta.
So was OpenDoc. And Newton.
Depends on the R&D cost it took to get there and the gross margin on that $3,500. Assuming Apple spent $150M on developing the thing (that number is probably low), and a gross margin of 35% (pretty typical for hardware), that tech firm would be $28M in the red. I do agree, though, that the AVP was never meant to be the end-goal of a product, it is a stepping stone on a larger roadmap, so calling it a "failure" at this point is very premature.
The Newton was a failure. The product line was killed 15 years before development on the iPhone even started. The only thing the Newton did that may have contributed to the iPhone's development was show other manufacturers such as Palm and Compaq that a PDA might be a saleable device, and the devices that were created by those companies in turn helped shape Apple's design direction on the iPhone. Inkwell was about the only Newton technology that even survived past the year 2000 (being added to Jaguar), but even that technology really didn't contribute much to the iPhone's development outside of generating industry speculation that Apple was working on a new Newton-inspired PDA of their own. As Blackberry had already proven by the time Apple started on the iPhone, even great handwriting recognition will lose out to mediocre keyboard input.