I think it's a mistake to view the Apple silicon mac pro as some big new investment. They had one very simple purpose - they said the whole Apple line up needed to be Apple silicon, and that's what they delivered.
The truth is that Intel spent a lot of time plugging away at all the different mechanisms that you could try to get performance, and one of those is high bandwidth peripheral interfaces. Apple hasn't got there yet, it doesn't have a chip with a tonne of transceivers on it that it can use to hook up a massive GPU, and to get there they need to not only get all the transceiver work sorted, they also need to build an entire software stack to deliver that performance. All to benefit the lowest revenue product in their line up chasing a sector that's not growing.
So what will happen? The same thing that happened with all of Apple's bad products- they'll continue to offer it for sale at incredibly high margins for eternity and never attempt to make it more competitive. Let me point you to the Mac Pro Trash can, it was almost immediately obvious that the product didn't deliver. What is the reasonable thing to do? Do a redesign as soon as possible! What did they do? Quote from Wikipedia:
>On September 18, 2018, the Mac Pro surpassed the Macintosh Plus's production life record for an unchanged Mac model, with the Plus having remained on sale unchanged for 1,734 days. It was discontinued on December 10, 2019, after being on sale unchanged for a record 2,182 days.
The new Mac Pro is going to be here for a very long time is my guess.
> So what will happen? The same thing that happened with all of Apple's bad products- they'll continue to offer it for sale at incredibly high margins for eternity and never attempt to make it more competitive.
There's another thing Apple do with bad products, though. The original 13" Aluminum Macbook launched in November 2008. It was discontinued in June 2009, in favour of the 13" Macbook Pro, which offered a permanently built-in battery (the previous model had a removable battery), and dropped Firewire. That stayed on sale, subject to regular spec bumps, until October 2012, when they bumped the screen to a retina display ...
I'm not an insider, just a long-term Apple watcher (been using Macs since 1991): what seems to happen is they used to get an internal goal to fill a niche in the product range by a certain date, and if the performance target isn't quite achievable they ship whatever they've got -- then iterate on it, possibly with a model branding change.
The Mac Pro looks like a classic case of this: Tim Cook committed Apple to being 100% Apple Silicon across the range by the end of 2022, they missed, but they had to ship something to keep the Mac Pro brand alive. So they shipped a turkey (a Mac Studio with PCI slots and a higher price tag).
What happens next is: either they ship the real Apple Silicon Mac Pro in a year or two (with who-knows-what changes, but something to differentiate it from the top of the Mac Studio range other than price), or they admit defeat and kill the Pro range altogether, or maybe downgrade it to be the name of the top end Mac Studio.
But this model won't be available for more than two years, max.
> What happens next is: either they ship the real Apple Silicon Mac Pro in a year or two (with who-knows-what changes, but something to differentiate it from the top of the Mac Studio range other than price), or they admit defeat and kill the Pro range altogether, or maybe downgrade it to be the name of the top end Mac Studio.
There were a lot of rumors around a quad-chip (a ultra is just 2 M* Max's) that was 2 ultras fused together but the yields were terrible and they scrapped it. I've heard speculation that they might be trying again with M2/M3's but yeah, as it stands now the Mac Pro is very lack luster.
Honestly even if they manage a quad-chip design it's still missing things professionals want (more ram, more GPU power). I'm not at all in the market for this device but I really hope Apple pulls it off. Apple is better when it has devices spanning the whole spectrum even if the edges (low-end phone, high-end desktop) don't have blockbuster sales. You sweet spot of sales is empowered by those 2 extremes and benefits from it.
I think they can solve for more RAM pretty easily with the M3. Apple silicon is pretty dedicated to having the GPU cores integrated with shared memory though so I don’t see them offering graphics card options.
An M3 extreme which is 2 or even 4 M3 ultras sounds like it might be doable. M3 is already going to be a huge performence step because it’s a new process though.
Just checked on MacTracker -- yes, the 5,5 Pro had firewire; it's immediate predecessor was the Macbook (no pro) 5,1, which didn't have FW but had the removable battery -- the 5,5 added Firewire.
Oh yeah, I'd forgotten about that aluminum Macbook (sans pro). I was working in a multimedia lab at the time and we were pretty dependent on firewire for student projects and putting the Macs into targeted disk mode.
I think it's really more simple than that - they have made architectural choices for Apple Silicon that simply don't fit the Mac Pro paradigm. Unified Memory Architecture primarily.
The Mac Pro lives on as a Mac Studio for people who need PCIe cards for their audio/video hardware to connect to. That's it.
I may have a poor understanding of the mechanisms involved but it doesn’t seem immediately obvious why unified memory should hold back the implementation of high bandwidth PCI-E. Except of course if they just haven’t consciously designed their processors with enough PCI-E capability
The only issue I see is their lack of support for video cards integrating their output into the thunderbolt ports. Which is a problem they pretty much solved in the last Mac Pro with their MPX extension to PCI-E, which conveniently provided additional power and allowed the injection of displayport signal into the thunderbolt outputs.
All of the problems with the Mac Pro are theoretically solvable. They could even add expandable memory if they really wanted to. But since the Mac Studio is so ridiculously fast, why bother? The number of people who are going to be held back by the Mac Studio is a minuscule market not worth serving, they would never make a profit on it, and ultimately Apple is a business not a charity.
The Mac Pro is Apple saying: “if you need more than 192GB of RAM, we can’t help. But if you need PCI-E, we have a lot of old Mac Pro enclosures left over from the past 3 years when we didn’t sell any Intel Mac Pros, here’s a Mac Studio in a bigger box, enjoy”
I completely agree - a computer like any system is a series of design choices, compromises and decisions. Apple's vision of the Mac appears to be of a small, svelte and powerful system with limited extensibility. These are the systems of which they sell the most.
Regarding socketable memory - I believe that their current memory setup is significantly faster than any system with socketed memory. Given that they market the memory bandwidth heavily in their specs I imagine this would lead to a perceived performance drop for what is essentially their most expensive system.
My one gripe with the new pro is that they have reintroduced a problem they themselves solved, because they chose to not support discrete GPUs. MPX looks like a really neat solution to both internal system cable management, and the routing of thunderbolt and display data through the system. It's a shame it only existed in one generation of one product
Interestingly enough from looking at pictures of the new system board - they have less power cable connectors than the 2019. Is the new system using a smaller capacity PSU than the old?
I think that is enough. Sometimes I want everything in one box. To be able to spec a computer with the expansion peripherals on the inside is very convenient.
The author of this Verge article had the same assumption, but couldn't find anyone in the VFX or video production business who uses Mac Pros like this.
The reasons given by the studio people interviewed make sense. They get a lot more graphics bang for the buck with Nvidia-equipped PCs, and they can upgrade piecemeal.
Additionally for video pros, they were burned by Apple's consumer-oriented FCP X product overhaul over a decade ago and the mistrust still lingers:
'And while Apple’s decision to overhaul Final Cut Pro was over a decade ago, studios still haven’t forgotten. Almost all of the video professionals I spoke to brought it up. “Every filmmaker in the world was using this,” Lebensfeld complained. “And they lost that whole market. They just don’t take it seriously.”
'“They really did screw us over on that,” Ford, the documentarian, agreed. “I was really upset.”
'I searched high and low and ended up connecting with over 20 professionals for this story in order to locate someone who enthusiastically wanted to buy the Mac Pro. I found exactly one: Drofa. He loves the cheese-grater design.
'“The killer feature is when somebody comes and says, ‘Okay, you have a Mac Pro,’ and I say, ‘Yeah, I can make a cheesesteak,’” he explained.'
Apple and others could just as well create an accelerator card with a bunch of M2s on it.
They could also create some sort of blade server.. they print their own cups in literally everything. Phone, laptops, headset, apple cinema displays. The production cost of an M1 is about $30.
Remember that the 5-year $1.5B contact with AWS will expire next year. They already have XCode cloud, and have enabled virtualization natively for a while
Either large Mac Pro based rack-servers or accelerators for rendering, transcoding, and AI/ML.
They could market those 10x$30 chips on a single board as something that is "10x" the power of the Mac Pro, at “only” $20000 a pop.
You can be very certain that NVDA did not get unnoticed, as Apple has the same capabilities, and sees a future in AI/ML.
Apple is not going to enter the commodity server business, that is just a race to the bottom and probably the reason they got out of it in the first place.
The only reason XCode Cloud (or XCode at all) exists is because people need it to develop for ios, if developers could choose anything else they would.
At this point they are a luxury consumer product company, that is where the margins are that have made them successful for the last 10 years.
I'm not oracle so I can't say if you are wrong but wouldn't that be a significant investment to go compete in a market where margins are not that great while risking to dilute their brand?
The thing is, the trash can Mac Pro kinda fits the market segment the Mac Studio is now in. If Apple had created the "Studio" range back then for it, rather than suggest it was the successor of the Pro, it would have been better received.
Apple don't tend to make the same mistake twice, this time they created the new marketing category for that product. 99% of people who would have been in the "Pro" category are happy getting a "Studio", plus a bunch upgrading from an iMac.
The Studio is the new Pro to them, it's just for marketing they needed to change the name and have a "continuation line" for the old product. As you said it's here for the long haul but mostly has a legacy range.
This is exactly me. I have a 2017 27' iMac that's really starting to feel sluggish. I've been holding out for a new iMac but I have a new Studio on the way in a couple of weeks.
The "trashcan" Mac Pro (late 2013) was my favourite Mac of all time - it actually fit my professional needs perfectly, and performed incredibly well in a small package for a reasonable price (I couldn't spec out a comparable Xeon system at the time that was cheaper). I don't think it could be considered a failure by any means. It was a sweet spot when it came to price/performance for many of us in tech/science industries, and perhaps that was the reason it wasn't replaced for a long time.
The Mac Pro is actually very competitive against similar $7,000 - $10,000 workstations; especially if you need a lot of high bandwidth memory.
However, it's not competitive with workstations that are above that because you'll quickly approach workstations that have Epyc and professional Nvidia cards.
We'll need to wait for an Mx "Extreme" SoC, if Apple ever makes one, to compete above the $7,000 price point.
I feel like I'd have a very hard time designing a $7000 system which doesn't completely blow the Mac Pro out of the water in terms of CPU and GPU performance. Do you have some examples of $7000 systems which are beaten by the $7000 Mac Pro's cut-down M2 Ultra in terms of CPU or GPU power?
EDIT: I don't get why this is getting downvoted. The M2 Ultra really doesn't have very impressive GPU power compared to even consumer offerings from nvidia and AMD. The competitive $7000 workstations would have to have something weaker than a 4080.
LLM inference is GPU bound and VRAM bound. Given quantization however, 2x3090 or 4090 (48GB) is enough VRAM to load 65B quantized llama derivatives (30-40GB). With exllama you can get maybe 20-30 tok/s with dual 4090s on ~1200W. With an M2 Ultra maybe 10 tok/s on ~178W.
(I don't have these devices but have been researching LLM inference performance in the interest of buying a machine for them.)
Advantage goes to M2 Ultra if you ever might need more than 48GB VRAM. I think it's unlikely nvidia is going to release a significantly higher VRAM consumer card anytime soon, since that's what their A100s are for.
Depends on what you do. Let's say Apple makes an M3 "Extreme" and sets max memory to 512GB of RAM.
That's 512GB of high bandwidth RAM that you'd have to piece together 7 Nvidia H100 80GB cards to match. Each H100 80GB card costs $30,000+. Seven of them would cost $210k alone.
Now of course 7 H100s would have more compute power, but if what you need is a lot of high bandwidth RAM such as a large LLM or highly detailed scene rendering, then maybe an M3 "Extreme" would have the best value out there, by far.
> So what will happen? The same thing that happened with all of Apple's bad products- they'll continue to offer it for sale at incredibly high margins for eternity and never attempt to make it more competitive.
That’s one of the good things about Apple- you know the product will stick around, rightly or wrongly. It’s not as bad now, but other big companies would try to moonshot with a new product then kill it if it didn’t get traction.
Things like the MS Surface and Google Pixel are the recent-ish exceptions to this where you could actually have confidence buying them as they are established product lines and they won’t get cancelled next month ( well you shouldn’t buy a pixel due to the chronic emergency call crash problems, but that’s a separate issue ).
This article confirms what some friends and I thought right at the time of announcement. You’re paying $3000 for some PCIe slots. And as others have pointed out, there’s very little demand for those slots.
It seemed like a given (to me, at least) that a new Mac Pro would have:
1. Slots
2. Storage expansion
3. The most CPU & GPU cores available in a Mac
4. DDR5 expansion, in addition to the SoC RAM
In fact, we got #1 and a little bit of #2 (two SATA ports). #3 is a miss, since the Studio offers the same configuration. #4 seems like the biggest miss of all.
Overall it just seems like a pointless machine. The number of use cases where a Mac Studio + TB4 enclosure(s) won’t do can’t possibly be large enough to justify keeping it in the lineup.
You’re paying $3000 for some PCIe slots. And as others have pointed out, there’s very little demand for those slots.
I think those two details are related in the opposite way people would expect. I think Apple knows extremely well that very few customers want those slots these days. That $3000 takes into account the considerable NRE costs that went into developing the Mac Pro since no other model has PCI slots.
This is designed to be a very low-volume product. One of the few “Pro” products Apple sells where the name is sincere.
MBP is sincerely named too (excepting the weird 13"), and with the same familiarity with their user base. Lots of actual professionals choose these models for their higher spec configurations, and have done so for several years despite increased expansion limitations.
Marc Gurman reported that Apple was making a 4x Max SoC that likely would have landed inside this Mac Pro.[0]
For whatever reason, this did not come true. I suspect that Apple ran into technical issues trying to combine 4x Max dies into one SoC and have all the CPUs, Neural Engines, and GPU cores act as a single chip to macOS. Another reason could be that Apple decided that it's uneconomical to design such an "extreme" SoC for a niche market/machine.
Personally, I think Apple will release an M3/M4 "Extreme" that will go into the Mac Pro. I also suspect that Apple will try to create some sort of "Apple Silicon Cloud" in order to expand the market for such monstrous chip. The traditional workstation is a declining market. Such cloud would also be a way for Apple to serve customers who want cloud workstations.
Apple's privacy focus might have driven them to imagine a market for private clouds. Run your company's custom AI on one or two of these monster towers and never need cloud AI again.
(I confess I don't know enough about cloud tech or AI to know how far off the mark I am; I'm just weaving threads out of reading headlines)
Whenever Apple does something kinda weird, always remember: Tim Cook is a supply chain guy.
Just like the current 13” MBP, just like the old iPhone SE, the “new” Mac Pro exists to fulfill some manufacturing contracts and/or use up existing stock of the chassis.
A “real” Apple Silicon Mac Pro will come in 2-5 years (my best guess is 3), mark my words.
Just out of curiosity, what would you expect from a 'real' pro that couldn't be done in the studio? It seems the integrated RAM has put constraints on how powerful they can create the machine (ie. multiple ultra chips).
The point of the pro, at least in the past, was the ability to build an incredibly powerful, expandable computer, relative to the other offerings at the time. But Apple has moved further toward integration to get wins in perf/efficiency. It seems to be a product that no longer fits in their ethos.
That you can get a machine that's just 2x as powerful as a laptop - at 2x the cost - isn't particularly compelling.
I super doubt that a low volume high end Mac is going to eschew the new unified memory architecture. Once Intel support leaves macOS, 100% of their kernel/gpu efforts across 100% of their devices will be based on an ARM SoC, so they can go on a code deletion/optimisation spree.
Investing in technology has an advantage of producing reusable tech down the line.
It is very difficult to predict which parts are going to be successful.
And if they manage to create a product definitively better that competitors (like the current gen ARM macbooks), they are going to sell a lot more than 50,000 per year. There are single companies which can afford to buy 50,000 units.
No idea, honestly. You're right, the product doesn't really make sense anymore. But at the same time, I can't see them simply dropping it. Maybe they'll surprise us.
They made the 14 and 16, but they presumably still had plenty of 13" frames left, so they're just selling both. Jobs-era Apple likely wouldn't have done that, as it causes the product lineup to be a good bit less clear, but in Cook-era Apple, that's an acceptable tradeoff.
There is some demand for those PCIe slots. Far fewer without GPUs, but there are definitely use cases where people will pay $3K for mostly IO (audio being an obvious case, network being a more niche one, unaccelerated video is one I’ve heard and don’t really understand but admittedly I’m very far from the target market). And of course the fairly obvious PCIe storage expansion use case, which I don’t think should be discounted (although it probably has been for this gen of MP).
The speculation I’ve seen (on Ars forums) about RAM is generally a consensus that this is a stopgap, with future base volume increases expected and a future multiple of die connections (a so called “Extreme” being a 4x Max). I don’t think RAM expansion is coming back to the Mac platform for the foreseeable future outside of some kind of hypothetical way to add more SoCs.
Just I/O I guess? I’m definitely not the target market so I’m just filling in blanks in my understanding of this as people more knowledgeable talk about the remaining use cases.
For $3000 I would expect a different SoC that has a second integrated memory controller with DDR5 support, even if it leads to an asymmetric memory configuration.
This is easy to say, but tape out costs for a reticle size slab of silicon on an interposer are ~10 figures, at least.
It makes zero economic sense for a low volume product like the Mac Pro, even with Apple margins.
EDIT: Also, they would have to seriously compromise performance. A number of GPU cores would have to be removed to make room for the bigger memory controller. The bus would have to be even wider with regular DDR5 DIMMs, and it would suck much more power.
Is that true even with a chiplet or north bridge approach? I feel like we’ve seen plenty of examples of CPUs laid out this way. Yeah with a performance penalty but that 192GB of L4 with direct GPU access would go a long way to making that better.
Apple could indeed fab a small memory controller chip, but I am not sure the M2 Ultra die has the interconnect to support such a thing? AMD makes this looks "easy," but it is not easy, and they built for Infinity Fabric from the start.
As said below, the answer is CXL. Apple would be stupid not to support CXL devices on the M3 or M4, though I would not be surprised if they skip it
The M2 series does not support CXL, as far as I know.
They would presumably use UDIMMs anyway, so maybe there is theoretically space?
Still, that would be one heck of a socket/motherboard. I can't imagine how many motherboard layers and socket pins they would need for 16+ channel DDR5 in such a tiny space... If its even possible to manufacture, it would be hilariously expensive.
It still wouldn't hurt to have as an option for, say, a swap disk in RAM instead of the SSD, or expose it as storage to macOS for use as scratch disks, or even just use it as a buffer/cache for the flash storage. There are all kinds of use cases where it would be beneficial to have a wicked-fast storage medium working between system storage and memory.
Just curious but wouldn’t a Mac Studio paired TB4 PCI-E expansion board/dock fill this same purpose?
Is there something inherently in the Apple ARM arch preventing this? I’m still on a Intel iMac where such a solution works a charm, so my natural upgrade path would be a Mac Studio and display.
Of course you are still limited to your thunderbolt bus but I guess TB4 is good enough for a lot of PCI-E use cases.
I know this is a meme, but if you're using these aren't you better off with a Windows/Linux box? You can shove an actual modern GPU in there as well and get considerably more peformance.
I struggle to think of workflows that require macOS for this, given it has conceeded the high end of performance to others.
The latest MKBHD video on this is interesting. You’re paying $3K for the Max Pro to have the PCIe ports. Almost the only difference per the Mac Studio.
USB-C displays only use the connector, the traffic then goes over alternate mode, mostly DisplayPort. For 4k@60, unless both devices support DP 1.4, then both USB3 pairs are used for DP and there's no wiring left for USB3, only for USB2.
It is almost the reverse of TB1/TB2, when a mini-DisplayPort connector was used to negotiate TB.
You can do MADI and Dante over thunderbolt or ethernet, and with the amount of compute UAD DSP days are definitely numbered (as shown by UA themselves now offering native). I don't see audio as a driver, given what I can do as a composer/producer on my laptop (m1pro)
For sure — the convergence towards Ethernet based AV protocols (esp. in broadcast) seems like it has a promising future (as we’ve seen with Dante), possibly for video too (SMPTE 2110, etc).
I'm not sure how you'd replace a decklink card with ethernet. The bandwidth just isn't there. Same issue with the TB ports, there's not nearly enough bandwidth for pro AV.
The interest in the Mac Pro is rather weird. It's an incredible niche product which should really only be covered by select publications for niche industries.
The "trash can" was the last Mac Pro that you should realistically consider to be at least partly targeted towards end consumers. The problem is that Apple won't come out and tell us exactly who the envision as the customers for the to last generations of the Mac Pro.
My best guess is that they either have NO IDEA who should buy this and only created it to shut up people on the internet and officially complete the ARM transition. Alternatively they have a few high profile customers who specifically asked for it and the "feature set" is tailored to those select customers.
People online, like myself, took a huge interest in the Mac Pro, not because we'd ever get one, at least not new, but because the prospect of a high power ARM motherboard with replaceable RAM, storage and PCI-E ports was just to enticing. We just wanted to see how they'd pull it of, and then they didn't.
Apple knows exactly who the 2019/2023 Mac Pros are for - people who will pay for the best specs and need PCIe cards.
The 2019 was a very flexible machine, the 2023 is a lot less flexible. Don't expect that to change - the GPU and RAM will be staying on the SoC because that's the architecture every other Apple device has. You really wouldn't want to be in a world where the lowest selling Mac is the only one that still uses NUMA.
Apple hasn’t supported Nvidia for years. Those people already switched a long time ago. Apple silicon is rather competitive with AMD, especially with all that unified RAM.
If your GPU workflow is the classic kind of parallel processing with a compute bottleneck, then a couple of discrete GPUs is going to be faster, for sure.
If your GPU workflow is instead VRAM bottlenecked with unpredictable data access patterns then Apple silicon has the potential to be faster with that crazy amount of uRAM.
There’s also a dark tail of applications currently using the CPU because CPU-GPU-CPU latency is too damn high. Having everything on the same SoC will be a substantial latency improvement, which will enable using the GPU in places that discrete GPUs can’t go.
So if you’re doing 3D/vfx you’re probably unhappy, although as far as I know none of these people use Macs anyway, they all use PCs with high-end Nvidia cards.
On the other hand, this chip is going to be incredible for audio, once applications start taking advantage of it. Audio programming is really latency sensitive so discrete GPUs are right out, but with an SoC you can start to rely on the GPU being able to serve your request on time.
But at least the did supported eGPU on intel max - maybe not Cuda but metal, OpenGL, OpenCL. Now eGPU is gone as well in Apple silicon even though we latest MacBook Pro supports TB 4.0
Before Apple Silicon, the performance ratio between a Mac Pro and the rest of their lineup was large. I think people were hoping the same ratio would somehow hold true today (i.e. rumors of a new M2 variant that was somehow two M2 Ultra stitched together, in the same way the Ultra is two Max, and the Max is two Pro...)
I'm sure someone out there is running a gigantic local storage setup with U.2 drives and the new Mac Pro, and is doing something that would require so many PCIe slots, but it does seem very niche.
There’s something about the 2019/2023 Mac Pro I never hear get brought up in these threads.
Yeah yeah they suck and are pointless and your Thinkpad dug out of a Goodwill dumpster running Arch Linux is 100x more powerful etc etc
But: they’re assembled in Austin, TX. Perhaps one of the only things Apple makes that’s assembled domestically? Are we sure the continued existence of the Mac Pro isn’t the result of some kind of tax scheme or trade tariff boondoggle?
And maybe also a trickle-down/tracer bullet for on-shoring?
If the world goes further mad and US-Chinese relations sour, and/or if the institutional racism of globalisation is finally dismantled, such that manufacturing costs a similar amount the world over, Apple already have a functioning PoC for US assembly.
Apple is just going to move manufacturing from China to India; in fact, they already are. Mac Pro can be assembled in the US because it’s a low volume high margin product. Manufacturing MacBooks, let alone iPhones in the US would still need to be bootstrapped from scratch, because the scale of Mac Pro production relative to the scale of other Apple production basically is scratch.
The saying goes that you need a third, stupidly expensive bottle of wine next to the cheap and the moderately expensive bottles of wine, if you want to boost sales of the moderately expensive bottle of wine.
Same thing for menu design. That $150 seafood tower??? That's just there to make the $47 steak ($5 extra for the au poivre sauce and $7 more for the creamed spinach) look like a deal. Sometimes people order the tower but it's just once or twice a night.
This may be the central reason why the Mac Pro still exists.
Ever since Jobs returned in 1997, Apple's Mac lineup has been designed around the idea of a ternary choice of "Good / Better / Best". They used to call the Power Mac configurations explicitly that.
I suppose putting the Mac Studio guts inside a Mac Pro case wasn't a big design change, as long as they have the Pro manufacturing line still open. By shipping this lackluster design, they can wringe out a bit more revenue from the investment made in the Mac Pro tooling.
This applies even to car manufacturers.
I seriously doubt the S-65 AMG from Mercedes sells in sufficient volumes to justify its existence on itself. But from a holistic brand perspective, Mercedes-Benz would be crazy to kill it.
The other way to look at this is that they failed to make the Mac Pro enough better than the Mac Studio or laptops to make it worthwhile. Graphics cards are one of the biggest use cases for PCIe slots and is something that Apple doesn't allow. If you are looking for the highest horsepower options, they are all Windows based. And I say this as someone who prefers macOS and definitely doesn't need this level of hardware.
That’s the symptom. The problem is just a trade off of the Apple Silicon architecture.
The M1/M2 is brilliant for the form factors that consist of about 90% of PC demand. Macs are better as a device than almost any PC on a few axes.
But… an architecture using an SoC to scale from an iPhone to a large personal server, which is what a modern era Workstation really is will hit limits.
IMO, the failure was they were boxed into dropping MacOS Intel support in the near future, and for whatever reason were unable to ship the Apple Silicon they intended to. So they shipped a compromised Mac Pro that meets the needs of a few key use cases that can sorta work with the limited top line memory.
Maybe - they may not have been able to package it. Workloads are usually CPU bound or memory bound. If I need 200 cores to get a TB of memory, that’s a waste of money.
Intel segmented their products because they had the similar constraints. There’s a mondo-expensive 8 core 4 GHz+ part, and a part with lots of cores at half the speed. In Intels case, the memory is of course not part of the SoC.
I have two Mac Studios. When I need "more GPU horsepower", I spin up a GPU VM for $0.40-$0.80/hour (depending on my needs) and work away. Then I shut it off.
I can't be the only one doing ML development this way.
I do have a Windows machine with an nVidia GPU (RTX 2070 Super), but I don't use it anymore (bought it years ago).
ML on the cloud is way more convenient because you can trivially adjust your cost based on what you're doing: Training? spin up something big/expensive. Inference? cheaper (less VRAM) is usually fine.
I also like that I can run multiple instances simultaneously, something that would be prohibitively expensive if I had to have multiple machines sitting around waiting for me to use them.
Craig Federighi in 2017: “I think if you use Xcode downloads as a metric, it’s possible software developers are actually our largest pro audience. It’s growing very quickly, it’s been fantastic.” https://daringfireball.net/2017/04/the_mac_pro_lives
I personally had a 2010 Mac Pro, and it was fantastic. But since then, Apple never again produced a Mac Pro that I wanted to or could buy. Going from $2500 base price in 2012 to $7000 in 2023 is absurd.
If you compare Apple's other Macs, for example the iMac had a base price of $1300 in 2012, and in 2023 it's... $1300.
The Mac Pro almost seems like a product that Apple didn't care about, but had to announce by WWDC to save face in their transition schedule to Apple Silicon.
I can visualize what could have happened 6 months before WWDC:
Tim Cook: So, where are we on the Mac Pro transition?
Hardware team: We've got nothing.
Tim Cook: We need to launch. What's the quickest way we can launch a Mac Pro by WWDC?
Hardware team: I guess we could slap an M2 Ultra on the motherboard of the last Mac Pro and call it a day.
> The Mac Pro almost seems like a product that Apple didn't care about, but had to announce by WWDC to save face in their transition schedule to Apple Silicon.
This is mostly correct. But it’s also mostly a product they haven’t cared about for approximately a decade. Two revisions, one completely rejected by almost all of its target market and positioned exactly opposite to the new offering. They know there’s a non-zero customer base they want to supply here, but every other part of their business is contrary to supplying it. The only reason they really care to supply it at all is network effects, and they also realize that’s way past diminishing returns. I think the only reason they even bothered besides prior commitments is lineage at this point.
It’s one of the highest profile products they make, and one of the longest running embarrassments too. If I were Tim Apple I’d have just cancelled the line quietly before the ISA transition, but it’s obvious that few thousand sales is more valuable to the company than the few millions in revenue.
I suspect that they ran into issues trying to combine 4x Max SoCs into a single SoC. Thus, they had no choice but to finish the transition by releasing a Mac Pro that uses an Ultra SoC rather than an "Extreme" SoC.
Marc Gurman received insider information about a 4x Max SoC 3 years ago. He was right about the 2x Max SoC which became the Ultra.
Reportedly they had yield issues at that size. Probably something they'll try to resolve for M3, or maybe release another M2 revision of the Mac Pro later.
Apple has some extremely wealthy customers that would just get the biggest and baddest mac they can get on principle. The Mac pro is an excellent way to tap into that customer base. The professional market for these is very niche but also very real. People running huge workloads in the media industry.
Apple doesn't make product decisions (in terms of design/arch) six months before an announcement. The lead time for most Apple products is closer to 2-3 years.
This is true. I'm a developer (Clojure mostly) and:
* I had a Mac Pro until 2016, sold it and used the money to buy a second monitor
* I had an iMac 2019-2022, sold it because I bought an M1 MacBook Pro 16", which was faster
Right now I very much enjoy a single-computer life, much less overhead. I anticipate I'll want to upgrade to an M3 MacBook Pro, but I don't see a need to use anything else, especially as the CPU (what I care about!) performance isn't significantly better in the Mac Studios or Mac Pros.
Indeed. I have an M2 Max MacBook Pro with 64GB and holy shit this thing just doesn't break a sweat for even heavy workload and is still small enough to be my every day laptop I can take with me to do basic work on when I travel about.
I used to need a big desktop/workstation for "real work" and had an ultrabook for all the related admin stuff that I like to do out of the office for sanity. Two machines isn't difficult to manage but it can be a bit annoying. Having one machine that can do everything without compromise is a dream.
Not to mention this one MacBook Pro, while expensive, is still cheaper than what I used to spend on a desktop+ultrabook setup. It is so close to perfect I pretty much have nothing to complain about which as my wife will attest to is unheard of for me when it come to computer :)
I had toyed with the idea of getting an Air and a Mac Studio and continuing to work in my old two machine way but with how good the MacBook Pro's are I realised I no longer needed to do that and I am very happy I made the decision to just go all in with a good MacBook Pro and not think about it anymore.
I have an M1 Max MBP w/ 64gb and I'm like you. Short of this thing breaking, I'm not sure what Apple can do to make me want to upgrade. It's very fast, does everything I need it to easily, has great battery life, etc...
Maybe if they add something like cellular to the MBP I might be tempted, but even then wifi is ubiquitous now.
I don't think I'll ever have another desktop computer, currently rocking a 14" M1 Pro with 32GB of Ram and I do FE web dev, backend dev (node/C#), some native Android/iOS dev, run docker containers, etc, etc
It just handles everything I throw at it with ease, I honestly don't think I'd benefit from upgrading to any of the M2s, or even a higher spec M1/more RAM.
I think it would be compatible, in the same way a GPU would be compatible. You wouldn't be able to use it for your display though (for the same reason you cant use integrated graphics with your GPU).. so it would only be useful for accelerating ML or graphics rendering workloads for example.
Domain-specific accelerators have always had their own memory. Look at GPUs. I don't see why a power user who needs speed would be put off by some DMA.
Maybe Apple ends up implementing something like CXL down the line, anyway.
Without the Apple proprietary MPX slots, such cards will be limited to 75W bus power. PCIe bus on M2 Ultra is also limited to something like PCIe 4.0 16+8 lanes, or 47.28Gbps(5.91GB/s) total. This will be much slower than competing(in price) product such as Intel Xeon W/AMD EPYC, which has up to 112/128 lanes of PCIe 5.0 or 441.28/504.32Gbps(55.16/63.04GB/s) max bandwidth.
This means a M2 Ultra PCIe card/backplane desktop supercomputer takes 192GB/5.9GB/s = 32.54s per node minimum to fill the whole 192GB of RAM - I don't know how to interpret this, is this going to mean the cluster will be I/O limited, or will it be actually limited by CPU?
I have an (almost) fully specced M2 Pro Mac Mini. Totally sufficient for all my music production needs (which does benefit from many cores). No reason to get a Mac Pro.
M2 pro mini owner here too. That thing rips through everything I throw at it.
I think a lot of the 'mac pro' thinking from 15 years ago was you needed a 'big pro machine' to do serious work. Nowadays, these machines only eat up space, energy, produce noise and while they can be faster the returns are diminishing and you don't need it for a lot of serious tasks.
It's a bit of the Unix workstation vs PC market repeat. While some people will hold on to their 'big PC' as long as they can Apple (and hopefully others) will release vertically integrated small machines that offer serious performance and eventually will take over the market.
I used to be all about the upgrades: my first main computer was a big tower PC, and I must've upgraded most of the components inside, including the power supply at one point. Always kept my just about 'on-trend' with minimal costs, but buggy as hell, especially under the version of Windows (98, etc.) around at the time.
I just replaced my 9-year old Macbook Pro with an M2 mini and, understandably, it feels like a massive upgrade. But the laptop didn't even feel particularly old after a decade, just 'underpowered'. I've realised that upgrading is a nice option to have, but it's even nicer just to not need it in the first place.
Yeah, in anything related to GPU power, Apple has lost the market, it is either Windows or GNU/Linux at this point, depending on which set of tools the companies are using.
That was always a reason why I only use Macs at the employer and not privately, the price tag for a gamer setup.
It is cheaper to get a graphics workstation laptop than one of these models, let alone a desktop model with a nice set of cards connected in SLI mode.
It's still probably a good product for big Audio/Video production houses to fill them up with PCIe cards for Video/Audio, super fast ethernet or SAN cards.
I appreciate that this product is geared towards a narrowly-defined type of user. But I'm genuinely curious why would it be more appealing than a Studio with a Thunderbolt enclosure?
PCIe configuration of the 2023 Mac Pro: 1
Two double-wide full length x16 gen4 slots
Two double-wide full length x8 gen4 slots
Two single-wide full length x8 gen4 slots
One single-wide half length x4 gen3 slot preconfigured with the Apple I/O card
and up to 192GB RAM.
Meanwhile, a typical real-world workstation mobo like the Asus Pro WS WRX80E-SAGE SE offers: 2
7 x PCIe 4.0/3.0 x16 slot(s)
Up to 2TB RAM
In my use case, I require 512GB RAM (next stop 1TB) and plenty of cores for heavy parallel build action, a Blue-team graphics card, 3 cards of fully-populated Asus Hyper M.2 X16 and a high-speed NIC, plus some life-improvements like a 4-port USB-C card. Requirements that seem to outstrip the capabilities of this new machine.
Hey, I love the appearance and feel, and I'm used-to and productive with MacOS on desktop. Up until iOS and newer cousins took over Apple's revenue sheet in the early 2010s, the x86 Mac Pro was a permanent fixture under my desk and a reliable partner. That plus a well-specced Linux box and I never had to worry about being out of action. But since that time, my impression is that it and Mac on the desktop in general feels like it's in decline, neglected, it's software increasingly abandonware, Apple's conception of the target market is every-more narrowly defined and on a just-enough basis, and perhaps most important of all the system's first allegiance no longer to me.
Maybe I need to accept or re-conceptualize that Mac Pro today is just for a narrow group of content creators, at least for the meantime...
That will tell you which tabs are problematic. For me it was a bit surprising, LinkedIn for some reason eats a bunch of CPU. Or maybe not so surprising, heh.
Then just unload the tab in tree style tab and check it when you need instead of having it always open.
I briefly considered a Mac Pro server rack for an ARM64 build server, but Asahi Linux doesn’t advertise M2 Pro support, and I didn’t want to fight with Docker on Mac disk speed woes (or other Mac/Linux compat issues). The Ampere ALTRA is currently my top candidate, it’s also much cheaper. :)
This is a very biased sample. Everyone who is still using Mac despite Apple giving NVIDIA the boot obviously isn't doing work where peak performance computing really matters - because that's just not a thing on Mac. Apple will never get there with their "big unified architecture" that's really a scaled up mobile chip, especially in terms of GPU.
Apple really has three options here:
- ignore that segment of the market
- create their own GPU AIBs and a software platform to make them attractive
- invite NVIDIA back onto the platform
Of these, the first one seems the most likely. The last one would be the most expedient, but conflicts with Apple policies against having a working solution that makes someone else look better than them.
Apple is focused on gaming, however only in what concerns their mobile devices, and for that, the current hardware is good enough for cross-device development.
"Exactly who these pros are and why the Mac Pro is the perfect device for them remains somewhat unclear to me, even after testing the new machine for a few days and speaking to various professionals that Apple is ostensibly targeting."
Yeah, because you're not one of them. Anyone that needs to run specialist PCI cards, and there are many people out there particularly in video capture (not just GPUs) will need to spend the money on this machine. And as for the oodles of power, high definition video will find a way to use it. We could have stopped at 4K but the quality and resolution chase isn't ever going to end because it tracks Moore's Law. Try being an artist using ultra-high def video for example, and try maxing out the machine for 6-12 months to see if you can get anything maximally creative out of it. Testing for a couple of days just means you're only using it for the basic journalistic stuff you're doing on every other Mac.
Just because you can't think of your use case for something, doesn't mean that nobody has a use for it.
Finally, if you're producing hundreds of thousands of dollars of value per year, then the ten grand cost of one of these machines is just a tax write-off. It's not designed as some competist's toy at that price, same with the phenomenally expensive display. Apple taxing money off the tax man.
The conceit of the article is that the author isnt a "pro" (whatever that means), so they went out and talked to a bunch of people in different industries, and no one had a use for the Mac Pro.
Tax loss. It’s not complex - capital expenditure is taken off the tax bill, so there’s people for whom these machines are phenomenally expensive and there’s people for whom they’re just a cost of doing business they’d be paying anyway. Apple know what they’re doing producing them.
Author talked to a handful of people, none of whom are using PCI cards or complex workflow. Even when he talks to the video editor who is using a MacBook for editing, he doesn’t consider that video has to be ingressed in a professional way, he’s just editing it. There are a bunch of use cases for this Mac Pro, and the author has committed a fallacy in saying “there are no use cases for the Mac Pro” just because he didn’t find anyone who needed one.
Yeah people can use Linux and windows workstations, sure. It would be trivial for Apple to abandon that sector of the market and stick to devices. I’m pretty sure that once the Apple Vision headset comes out, and perhaps a generation or two down the line with wholly-Apple chipsets, certain people will be clamouring for Mac Pros again (i.e. - stereo feeds from phenomenally high definition video cameras to provide content real-time…). But by all means stick with your shade throwing and sarcasm, what do I know, I just have 25 years experience in the industry.
> The former, nevertheless, costs at least $3,000 more
But why? I would think with less size, energy, and weight constraints, it would be cheaper to make. Or at least it could be cheaper. And if it cost less, there would be a reason to choose it. So the only reasons I can think of for it to be so much more expensive, is they don't want it to compete with MacBook sales, and/or they are trying to extract as much revenue as possible from the niche customers who absolutely need PCIe slots.
I think this is unlikely. All of Apple Silicon is based on a unified memory architecture. GPUs and CPUs share the same, incredibly high-bandwidth memory. In fact, the GPUs CPUs in unified memory are all part of the same chip package.
Whereas All current PCI-E graphics cards include their memory on the card itself. Which means the CPU is having to copy data to, and from the external graphics card to do basic video operations.
In this interview on The Talk Show, John Ternus, vice president of Apple hardware, says he sees no way to incorporate external graphic cards into the Apple Silicon architecture (at 22:48)
Do they? While eGPUs were a thing for Intel macs, I don’t know if any external GPUs they could be used for graphics with Apple silicon, even in a thunderbolt enclosure.
My guess is it would be possible to get Asahi Linux to interface with a card, but it would likely run into similar quirks that we hit on other Arm platforms (like AMD driver issues with cache coherence on Ampere).
IIRC the actual limitation was the M1 only allows mapping PCIE as Device-nGnRE and that precludes eGPU/video cards. Using Linux doesn’t change this behavior.
Disclaimer: completely unaware of what that means.
Does it preclude such a card working _at all_ or can it still be used e.g for e.g CUDA?
Would it be possible for e.g Nvidia to create a special batch of cards or design that would work for such cases? (priced through the roof given the small user base but that'd also be the kind of user base who would drop stupid amounts of money without batting an eye)
Considering nVidia has the Grace CPU and probably a significant investment in arm64 development, I'd say yes absolutely. If anyone could do it it's them. Of course theres probably a better chance of Atlantis rising from the sea.
>> This isn’t just for pros, Apple seems to claim; it’s for capital-P Pros.
paying $15K for a Mac Pro (without screen) to be tethered to a desk isn't a very creative solution in 2023, WFH / remote working requires agility. MacBook Pros are very powerful already.
Plus, consumers are going to want more Energy Efficient / Green products, and that means long-lasting too (reduce waste).
I really enjoy the luxury of both. I have a MacBook Air that I use when I'm traveling or sick of being at the desk, but 80% of my computing is done at my desk with my Mac Studio. git, icloud, airplay and basically all corporate work happening in office360 means that I'm not really tethered to any one machine.
Why have the desktop? Power. M1 Ultra is a beast, throwing a little more money at a device that isn't going to die because the battery expired or I crack the screen made sense to me. I'm also mostly at my desk, mostly for ergonomic reasons: I hurt _less_ if I'm in a good chair where everything is tuned to my height.
Oh dear, don't tell me about the cracked screen - damaged two in 3 years :(
I like the energy efficiency of apple silicon chips, I think they've hit jackpot with that.
True, you get raw power from a high-end desktop/workstation, but cloud services have eaten-up missive chunks of of dev work. MacBook is a good medium (Jack of all trades)
I agree. demand for richer media Content is only going to grow, wouldn't be surprised if post-prod and the like didn't become the sole focus of Mac Pro over the next few years.
Like others mentioned in this thread, I don't think this is Apple's final vision for a Apple Silicon Mac Pro. To me, the only way forward is for Apple to, either use PCIe, or another proprietary high bandwidth bus solution, to connect together multiple Apple Silicon boards. This would result in a Mac Pro in which you can slide in multiple Apple Silicon modules, i.e. Mx Ultra modules.
For distributed Machine Learning workloads it also make sense to combine a CPU and a GPU per pluggable module. In such workloads, data that needs to be processed by a specific GPU (i.e. for training), is usually preprocessed by a CPU. In Intel machines with multiple Nvidia cards, this implies that you need a single beefy CPU with high core count. Having a CPU and GPU in each module makes sense from this perspective.
I've owned every Mac Pro with the exception of the 2019. I bought them because I wanted to be able to easily expand RAM and storage without having to use external enclosures. Apple doesn't seem able or willing to allow this anymore on their systems since everything is either part of the SoC, or hardwired to the system board. The new Mac Pro does allow you to add either NVME chips (at an incredible cost) or NVME cards in the PCI slots.
So as soon as the new Mac Pro was announced at WWDC, I did a quick check on the diff between it and the Mac Studio and instantly bought the Studio. I've had it for 2 weeks and it's been fantastic. Although I loved the old flexibility, part of that's gone, and part of it is just not worth the price Apple wants. And Apple has just pushed this price way above the value. Now that the performance is the same, they've limited the market dramatically.
Actually, the Mac Pro’s biggest problem is the iPhone. Or rather, Apple’s one-trick-pony design of their quote unquote "desktop class" SoC to be just the iPhone SoC, but a little bigger, with the only scale options being multiples of the same chip, with everything soldered on the SoC, and then some thickheadedness on “we didn’t design the architecture with external GPUs in mind”. Apple is already getting beaten by performance-oriented traditional CPU systems, and soon, other vendors will optimize for battery life as well. And Apple will still be stuck with an iPhone chip in all their computers.
But that’s OK, because it seems they are pushing their desktop OS in the direction of a phone OS too, both in capability and lockdown, so it might all be good.
Having just upgraded from Mojave to Ventura, I don't really see that they're heavily pushing the desktop OS in the direction of a phone OS.
This is still the most powerful Unix desktop system out there, and there's nothing keeping "power users" (in lack of a better term) from treating it as such.
They've been talking about this for years though, and I think with people like Craig Federighi in place, there's no real danger of further lock downs.
Just take a look at the security hardening at disk level, war on kernel extensions, continued reliance on entitlements for basic functionality (that is not just limited to sandboxed processes). Coupled with bootargs (like amfi_get_out_of_my_way) being broken for multiple OS versions, and it leads me to a pretty clear conclusion.
I did a piece about four years ago where I tried to determine who was still willing to buy the trash can, new, in 2019, and found that the answer was (essentially) MacStadium.[1]
This piece, while quite good, doesn’t dive into this, but I am curious whether MacStadium has plans to do anything with the Mac Pros that they wouldn’t with the Studios.
$3,000 for access to slots that you can get on a $500 desktop seems like an upcharge designed to discourage people from buying it.
But the reason mac pros had PCIe slots was mainly for graphics cards.
and admittedly graphics cards are now more than graphics cards now, they do GPU accelerated computing and machine learning. And apple software would not support those things on a GPU.
It is sort of like Steven Covey's progression from Dependence to Independence to Interdependence. Apple with the early intel mac pro broke through from independence to interdependence. That machine used 3rd-party intel processors, worked with not only macos, but 3rd party windows, had expandability via memory and 3rd party hard drives, and used 3rd party graphics cards from nvidia and amd. They worked with the entire silicon valley ecosystem, making their machines more valuable and the ecosystem stronger.
Now apple is independent again, and is looking right down into its own navel like a self-involved loner.
It was part of the reason. Definitely not the only reason, though. Do you think super fast networking cards or super fast storage cards aren't worth anything?
Anyone old enough to remember the ways of Apple before they almost closed shop can reckonize that in a way they are back to their old ways, now that they are ridding the wave.
one of the main reasons the mac pro had PCIe slots was for graphics cards. now, if that was the only reason you needed a mac pro you can get a mac studio instead for much less money. lucky you. it's not as if apple didn't expect this, or it's a problem. they created the product, pretty clearly because they knew people would want it.
the mac pro only exists for the few use cases where you need a non-gfx pcie card. if you don't have one of those use cases, then you don't need a mac pro and apple has lots of other computers they will be happy to sell you. theverge article seems to think this is somehow bad?
Might lean toward postproduction studios' use cases?
No GPU support, but the slide during the presentation showed extended storage (4-8xNVME to PCIe adapters are fairly inexpensive I think), SDI video ingestion, ProTools HDX cards (music I/O and DSP), 25GbE card, and I think a fiber channel one. Since then a third-party company announced a networked audio card.
Possibly some special-purpose accelerator cards; IIRC the first-party Afterburner cards on the last mac pro were FPGA-based; video acceleration is now built into the M2 chips.
I think most of the extra ports that differentiate a Mac Pro 2023 from a Mac Studio Ultra on the spec sheet are on a small PCIe card included with the machine that stays in the top slot.
a) Professionals who have a lot of external TB appliances e.g. audio interfaces and want everything in one case. To save on the hassle when moving between jobs or just to simplify their studio.
I think the lack of expandable memory is worse than lack of support for dGPUs. Everyone doing technical work (e.g. scientific computing and audio/video and VFX) could use more memory, but not everyone cares about GPU performance. They need to figure out the expandable memory piece first. This would instantly open up the scientific computing market to the new Mac Pro.
I think it’s a matter of lifestyle choices as well when comparing to the studio. A lot of us work on our computers for a good portion of the day, why not have the $3000 upgrade that looks way cooler?
People make these upgrades in trim options on their cars all the time.
Apple went out of their way to add the ability to boot third party operating systems to M-series macs, which the Asahi Linux team has been using to build a Linux distro tailored to M-series Macs.
Asahi Linux is still a work in progress but shockingly usable for its age, and will no doubt be ready to roll by the time the first M-series macs have their support dropped. I’ve read that OpenBSD is bootable on M-series macs too, so that may also be an option.
Additionally, while I don’t think Microsoft will ever take the invitation, Apple has publicly stated they’d welcome an Apple Silicon port of Windows.
I'm never going to get a pro, but I've recently found the thermal performance of the studio to be disappointing compared to its size. It doesn't take much CPU numbercrunching to slow down to a crawl.
The truth is that Intel spent a lot of time plugging away at all the different mechanisms that you could try to get performance, and one of those is high bandwidth peripheral interfaces. Apple hasn't got there yet, it doesn't have a chip with a tonne of transceivers on it that it can use to hook up a massive GPU, and to get there they need to not only get all the transceiver work sorted, they also need to build an entire software stack to deliver that performance. All to benefit the lowest revenue product in their line up chasing a sector that's not growing.
So what will happen? The same thing that happened with all of Apple's bad products- they'll continue to offer it for sale at incredibly high margins for eternity and never attempt to make it more competitive. Let me point you to the Mac Pro Trash can, it was almost immediately obvious that the product didn't deliver. What is the reasonable thing to do? Do a redesign as soon as possible! What did they do? Quote from Wikipedia:
>On September 18, 2018, the Mac Pro surpassed the Macintosh Plus's production life record for an unchanged Mac model, with the Plus having remained on sale unchanged for 1,734 days. It was discontinued on December 10, 2019, after being on sale unchanged for a record 2,182 days.
The new Mac Pro is going to be here for a very long time is my guess.