MP All Models Intel may be the only way to save the Mac Pro

[prefuse07]

This whole thing is very interesting.

On one hand, there's the possibility that Apple backed themselves into a corner--meaning, they didn't anticipate the extent to which Apple Silicon doesn't work with the Mac Pro's modular/upgradeable philosophy and now they don't know what to do. I find that hard to believe that they couldn't have seen this coming.

Well, wasn't Timmy the one who said something along the lines of "all people really need is an iPad"...?... That should give an indication of how he thinks, and the direction to which he is currently driving the mothership...

The other interesting thing to think about is this:

1. The Mac Pro was developed during the Steve Jobs era, and look at how amazing it is as a piece of engineering...

2. The Mac Studio was developed during the Timmy boy era, and look at how limited and constrained it is...

At the same time, rumors indicating the Apple Silicon Mac Pro will be minimally upgradeable and will have the same chip as the Mac Studio lends some credence to this theory. Right now it seems like the Mac Pro is going to be a Mac Studio in a different form factor will possible upgradeable storage. Is that enough to differentiate it from the Studio? Why keep the Mac Pro around if it doesn't offer much over the Studio? (Or is the theory that the Studio is a "stop gap" that will be quickly discontinued once the new Mac Pro is released true?)

I am of the mindset that the Studio is indeed a "stop gap", one of which they quickly scrambled to put together, and didn't give much thought to, and I strongly believe it will be discontinued once the AS Mac Pro, and newer AS chips are released.

Have we already seen the limits of Apple Silicon with the Studio?

In terms of GPU power, the Studio already gets easily smoked by an RX-6800XT (in a 5,1 at that)

So take from that, what you will...

At this point I just want to see where this goes.

I am with you 100% on this.

I am even more excited to see the direction apple will take once Timmy boy finally resigns. Hopefully we get another Steve Jobs type character, that actually pushes the envelope, because with the amount of cash that they are sitting on, it really is quite a waste to not use any of that to innovate.

 

[subjonas]

Hopefully we get another Steve Jobs type character,

Sadly, I don’t think they come around all that often.

 

[steve123]

Apple Silicon doesn't work with the Mac Pro's modular/upgradeable philosophy

There is nothing about AS that does not lend itself well to a modular philosophy. In fact, the existing Mac Pro tells us much about what the new Mac Pro will look like. Apple will likely offer MPX modules with up to 4 x M3 Max chips and up to 392 GB RAM. Each MPX module will have 4 TB5 ports and PCIe Gen 5. You will have a great deal of configurability by selecting the modules you want to plug into the machine. You want more RAM, upgrade by buying an MPX module with more RAM. You need more cores, get an MPX module with more cores. You want more GPU's, get an MPX module with more GPU's. Moreover, when M4 comes along, you will be able to mix and match. add an M4 MPX module to your machine without having to buy an entirely new machine. It is going to be a killer machine. There will be nothing like it for years.

The tower will have slots for up to 4 MPX modules as well as several PCIe Gen 5 slots.

 

[mattspace]

I am of the mindset that the Studio is indeed a "stop gap", one of which they quickly scrambled to put together, and didn't give much thought to, and I strongly believe it will be discontinued once the AS Mac Pro, and newer AS chips are released.

That's where we have a difference of perspective on it - because while like most people here, I'm not really interested in a Studio, I don't think it's a bad Apple product per-se. I think a Mac Studio, plus a Studio Display is a better product than a 27" iMac. It's more expensive, which is not great, but it also allows your monitor investment to span computer generations, without the kludge of Target Display Mode.

I think it's a very clever, calculated product whose purpose is to replace big-screen iMacs permanently. It's a strategy product - the new strategy for the desktop is "modularity" (that terrible interpretation meaning "computer and monitor are separate"), and the AIO of the iMac is going to be the outlier.

Th Studio is better in terms of:

• Lowering eWaste and embodied energy loss.
• Reducing SKUs, while increasing the breadth of configurations possible. *edit* Think about retail store stock management when all of the largest, most space consuming boxes are just the one product (a display) vs having to devote space to multiple big iMac configs.
• Increasing cashflow with more regular, cheaper computer upgrades between rarer monitor sales.
• Reducing average age of the install-base.

I think a lot of the idea of the Studio as being a fill-in product, or scrambled to be put together is driven by a worldview of Apple that is 27" iMac-centric - as if that computer and form factor is somehow the anchor of the range. That worldview relies on an assumption that an AS 27" iMac is in the works, but there's no reason to think that because there's no reason for the 27" iMac to be held back if it's a product still in development.

The iMac was the cheap, small computer from the start (hence the 24" iMac being the only model now), and Apple is deeply into nostalgically ouroborosing its own history, in lieu of having any genuinely new ideas.

 

[steve123]

Lowering eWaste

That is a good point ... likely why the 27" iMac will never return.

In my view, the Studio is the evolution of the Trash Can. The Trash Can was an anomaly in the MP line up but it definitely has a loyal following.

 

[singhs.apps]

That's where we have a difference of perspective on it - because while like most people here, I'm not really interested in a Studio, I don't think it's a bad Apple product per-se. I think a Mac Studio, plus a Studio Display is a better product than a 27" iMac. It's more expensive, which is not great, but it also allows your monitor investment to span computer generations, without the kludge of Target Display Mode.

I think it's a very clever, calculated product whose purpose is to replace big-screen iMacs permanently. It's a strategy product - the new strategy for the desktop is "modularity" (that terrible interpretation meaning "computer and monitor are separate"), and the AIO of the iMac is going to be the outlier.

Th Studio is better in terms of:

• Lowering eWaste and embodied energy loss.
• Reducing SKUs, while increasing the breadth of configurations possible. *edit* Think about retail store stock management when all of the largest, most space consuming boxes are just the one product (a display) vs having to devote space to multiple big iMac configs.
• Increasing cashflow with more regular, cheaper computer upgrades between rarer monitor sales.
• Reducing average age of the install-base.

I think a lot of the idea of the Studio as being a fill-in product, or scrambled to be put together is driven by a worldview of Apple that is 27" iMac-centric - as if that computer and form factor is somehow the anchor of the range. That worldview relies on an assumption that an AS 27" iMac is in the works, but there's no reason to think that because there's no reason for the 27" iMac to be held back if it's a product still in development.

The iMac was the cheap, small computer from the start (hence the 24" iMac being the only model now), and Apple is deeply into nostalgically ouroborosing its own history, in lieu of having any genuinely new ideas.

Agree with most of what you say, but I think a touchscreen Mac will surface (pun intended) in due course of time.
If such a thing happens, it’ll be interesting whether Apple will go for pure touch+pencil screen or create an iMac around those lines (or both).

The iPad is very good, but I find the screen a tad too small sometimes. The portability aspect of the iPad limits the screen size.

The Wacom market is waiting to be swallowed by Apple, Microsoft, Samsung etc tiny as it is.

Back to topic.
Yes. The mini+studio stack will most like mirror the Mac book lineup minus extra fluff (screen, keyboard, speaker, touch pad, battery, webcam etc). This will end up making it cheaper than MacBooks. Perfect for peeps who do their work on the desktop.

 

[ThisBougieLife]

I have to wonder if another reason the 27" iMac was canned was because they weren't able to get a larger panel. It's telling that the Studio Display is 27". They increased the size of the smaller iMac, but not the larger iMac's replacement.

 

[ChrisA]

I think we recently got a solid clue about what Apple is doing. A new macOS release has the words "compute module" in the code. In the computer industry, a "compute module" is a little printed circuit with a CPU and some other support chips and RAM. They are usually sold to industrial users who place them inside products like robots, cars or whatever. They are computers that lack parts a human would use like displays and keyboards USB or storage

So we know Apple is working on a compute module. I suspect these will go into a future Mac Pro. Maybe each module has an M2 Pro chip with 64GB RAM and 12 CPU cores. Then you buy the Mac Pro and you get to choose how many compute modules the Mac has inside. Later you can upgrade the modules with the M4 is released

The Mac Pro would then be like a network with many Mac Studio computers connected by a VERY FAST kind of network. This is how modern super computers are designed.

Depending on the bus that these modules plug into, there could be any number of modules installed on a bus. Then in a few years a new a Mac Pro "Ultra" might have multiple buses. The idea could scale up to a data center size.

The first Mac Pro might just have one module, but at least you would be able to upgrade it to an M4 later (I assume the first Mac Pro will have the M3.)

The problem is the software. You have to write it so that it can distribute the work to all the networked modules.

 

[mattspace]

Agree with most of what you say, but I think a touchscreen Mac will surface (pun intended) in due course of time.

I think they're most of the way there already with the ability to use an iPad as a display. The biggest weakness of the Surface Studio is that it's an AIO*. I'd be unsurprised to see Apple just do larger and larger iPads, limited by the bandwidth of Thunderbolt, that can be a display when wired to a Mac.

*But ho boy, would I like to try a workstation with a Surface Studio, flanked by twin portrait oriented touch displays, for palettes, media choosers etc, keeping the Studio as a workbench.

I recall a video that I haven't been able to find since, of someone using a surface studio for interior design, where they had a projector / biiiig tv on the wall, for the viewport, then using the surface dial on the surface studio to rotate the viewport & move about the space, while painting on the surface's screen with the pen to drop in colour / material swatches etc.

Absolutely beautiful looking work process.

*Edit* one thing I think may be a problem for the Mac Studio - I have to wonder if Apple's eventually going to encounter a trademark dispute with Microsoft over the name.

 

[singhs.apps]

Edit* one thing I think may be a problem for the Mac Studio - I have to wonder if Apple's eventually going to encounter a trademark dispute with Microsoft over the name

Not sure about Studio being an issue. Both are preceded by respective trademarks, in my limited understanding.

And yes. A 27/32” iMac (with ultra class Soc at the top end ) or a 32” 4k/6k screen with pencil + touch will be an amazing product (I prefer the latter)

 

[allan.nyholm]

If a new supposed Intel Mac Pro is coming then Apple would have to reverse their stance on some of macOS' feature set that is Apple Silicon and not Intel bases Macs. There's something about a Maps feature that I thing is related to the 3D view(unsure comment because I don't care enough right now to look)

And possibly also reverse their god awful stance on Macs allowed and capable of both running macOS Ventura and include all the feature sets to match their Apple Silicon macOS Ventura feature set minus the obvious? iOS app launching.

We can't have a top modern new Intel Mac Pro only capable of changing the wallpaper.

 

[Boil]

Much like surviving an asteroid field, the odds of a 8,1 Intel Mac Pro are 3,720 to 1...

 

[opeter]

You mean like this ?

Was a stupid idea 37 years ago is still one today, even before going into the cost and performance issues anything like this or with tighter integration would cause.

Well, there was this too:

Power Macintosh 6100 - Wikipedia

en.wikipedia.org

Here is a video with in-depth analysis:

 

[dazzer21-2]

The last opinion piece I saw implied that the Mac Pro is pretty much a chassis and that's it - the guts of the machine will come as swappable modules. My take on that is that slots will be filled with the equivalent of what are now MPX modules, each containing CPU, GPU, RAM and expansion - in other words, a Mac Studio PCB/SOC on a 'tray', rather than utilising single swappable items each for CPU, GPU, RAM etc. I'd assume, then, that Apple have found a way that multiple trays in this chassis can work in parallel. So taking current top tier Mac Studio as a base, two trays effectively add up to 40C CPU, 128C GPU (would the neural engine count double up?), 256GB RAM, 16TB storage and the associated ports (SD card slots on the front, anyone? 😜). Plus proper cooling. If the Studio's current $8k per tray holds, to build that system up to the same price as the current top tier 1.5TB Intel Mac Pro, that should - providing these trays can communicate at full speed - make up for quite the powerhouse. Maybe they'll upgrade the potential specs of these trays at the same time and take things even higher. Four slots sounds plausible, so that could effectively double those figures again. The trays, I would also imagine, could be fairly low profile in size so maybe in the current Mac Pro case, even more slots/trays could be a consideration. But what, other than future expandability, would the Mac Pro offer over the similarly spec'd Studio? Would they kill that off, maybe? It all sounds a bit too fantastic and bordering on improbability, but think of the potential if this was the way forward... If the Studio were to be binned, that ought to offer hope of a new 27" 5K iMac to fill the void that would otherwise be left for the mini Pro (Max?) and the Studio Display. Watching with interest.

 

[tsialex]

one thing I think may be a problem for the Mac Studio - I have to wonder if Apple's eventually going to encounter a trademark dispute with Microsoft over the name.

Mac Studio while is the first Mac that have the Studio name/brand that I know of, is not the first Apple product that have the Studio name/brand, a whole line of displays was called Studio back in the end of 90's early 2000's and Apple have the Logic/DVD/Final Cut Studio trademarks for eons.

I doubt that Microsoft can have any foothold of the name Studio alone (without Surface) or even is possible to trademark it generically to computers or apps, there are several other apps with Studio in the name, like Sound Studio, since the early PowerPC days.

Look how much products have the Studio name:

studio Trademarks - Justia Trademarks

trademarks.justia.com

 

[mattspace]

Mac Studio while is the first Mac that have the Studio name/brand that I know of, is not the first Apple product that have the Studio name/brand

Yeah, my thoughts were specific to the Mac Studio though, in the context of Apple producing a touch enabled display we were discussing - how close that would make a "Mac" Studio to a "Surface" Studio, and whether Microsoft could argue it's capable of causing customer confusion, or dilution of the distinctiveness of their machine's branding. Trademarks can be quite specific to the thing they're applied to - a mark for software would not also cover hardware etc.

 

[Unregistered 4U]

Have we already seen the limits of Apple Silicon with the Studio?

No, in pretty much every case, regardless of what Apple provides, we’re looking at code that has been honed and optimized for many years more than Apple’s Silicon has even existed. Developers still have a ways to go before they really have a handle on how to best handle Apple’s architecture. There ARE those that are leaning into the strengths of Apple Silicon and currently doing things not possible on anything produced at any price by Intel/AMD/Nvidia. That kind of knowledge is going to expand slowly but inexorably to the larger developer community and performance, even on older Apple Silicon systems, will improve as a result.

 

[Basic75]

No, in pretty much every case, regardless of what Apple provides, we’re looking at code that has been honed and optimized for many years more than Apple’s Silicon has even existed. Developers still have a ways to go before they really have a handle on how to best handle Apple’s architecture. There ARE those that are leaning into the strengths of Apple Silicon and currently doing things not possible on anything produced at any price by Intel/AMD/Nvidia. That kind of knowledge is going to expand slowly but inexorably to the larger developer community and performance, even on older Apple Silicon systems, will improve as a result.

What kind of optimisations are you talking about? Because in my book "optimised for" is just marketing speak, and most developers don't have the time or money to go beyond using swiftc with optimisations enabled. I'm quite sure they are not going through the assembly output looking for optimisation possibilities. Of course there are higher level optimisations, but which of those are specific to ARM processors and Apple Silicon? Sure, the accelerate framework will be heavily optimised for the new architecture, but that was available on Intel, too, so applications will automatically benefit from day 1. So what kind of optimisations are you referring to?

 

[Bug-Creator]

I'm quite sure they are not going through the assembly output looking for optimisation possibilities.

Depends....

There are various 3rd party libraries being used by most apps that might not have been fully updated
With modern CPUs and the same job done can be done in several different way, have it all run on the CPU the GPU or maybe even with specialized HW. Split it into parallel threads? Sure, how many? All doing the same on a part of the data or a more cooperative style?
Once all that is fixed you might even look at small (and heavily used ) loops to see if can make them fit better into the closest cache or avoid the CPU from staling on a wrong branch predictions.

Only makes sense for stuff that is really performance sensitive.