Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Mac Pro and Pro Display XDR Launched Three Years Ago Today
- Thread starter MacRumors
- Start date
- Sort by reaction score
Posters in here are professionals, and are representative despite your attempt to discredit them as otherwise.
Hate to tell you but IT professionals don’t go through all their training, and qualifications to become pen pushers for leasing companies. That’s a small part of it. If you‘re talking big business then they are usually departments that means budgets for each department, and you think they’ll be able to justify a Mac computer with say 2 grands worth of memory upgrades over something else.
Companies will maybe lease a laptop, a bog standard desktop, but a pro level high end workstation? It seems many claim a lot of companies moved away from the Apple platform when the trash can was introduced, it’s why Apple then launched the current 100% upgradable Mac Pro, so even Apples own actions and direction would suggest your observation is wrong.
I suspect if Apple moved to an even more locked down machine with zero upgrade options, More would leave the platform, not lease them.
Only if you take "landfill" completely literally as opposed to "dumped, buried or incinerated".
The aluminium case is nigh-on 100% recyclable, and people have been profitably "recycling" scrap metal since before it was called recycling. If I just dropped an old Mac off at the local
The problematic "e-waste" of the actual innards might be worth processing (at significant expense) but the output from that is going to be a few grams of copper, maybe a tiny speck of gold and a big heap of contaminated, mixed up plastic, glass and ceramics that still have to be dumped, buried or incinerated.
My question is - assuming you make reasonable 'proper disposal' efforts - how does the volume of actual lifetime e-waste (ignoring the easily recyclable case) generated by a modular system that gets its GPU and RAM upgraded a few times compare with 2-3 new Mac Studio-type systems. ...and I really don't think the innards of a Mac Studio contain significantly more e-waste than a single mid/high-end PCIe GPU.
...which is why there are incentives to recycle. Also, it's usually easier to find someone who wants a hand-me-down 2-3 year old complete & working computer than someone who wants to "upgrade" to the old GPU/RAM/SSD that you've just thrown out.
NB - I'm not saying it's a night and day thing, either for environmental or technical reasons: in the past I'd have taken the modular route but I think the balance is shifting with Apple Silicon turning system-on-a-chip architecture into a performance advantage - and I expect the rest of the industry will follow a similar route.
If it's LPDDR (low power) RAM it has to be soldered in. Put very crudely - the faster the data rate, the more current you have to pump down wires but that can be minimised by making the wires as short and unobstructed as possible.
In the past, Apple has sometimes soldered in RAM just for planned obsolescence and unnecessary thin-ness, but since MacBooks went to low power LPDDR RAM - and especially since Apple Silicon went with RAM chips mounted directly on the processor package - it's non-upgradeable for good technical reasons.
To have socketed RAM (a) Apple would have to build chips with a regular DDR5 memory controller and (b) take either a speed or power consumption hit.
They could stand to make the larger RAM configurations a bit cheaper, and be more generous with the base model, though. There's no reason to be making 8GB systems in 2022.
I am VERY interested to see what Apple does with this. The current Mac Pro supports 1.5TB of RAM, which is much larger than any current M1/M2 product.
Offering a new computer with a lower maximum RAM than the previous Mac Pro seems ... unlikely. But if the RAM is integrated into the SoC, offering choices from 64GB - 1.5TB would be a HUGE number of options and an inventory nightmare of producing all those different boards. (Aside from the fact that 1.5TB of RAM from Apple would be an enormous rip-off)
Well, they just have, in the form of the Mac Studio - and, 2 years in to Apple Silicon, there's no official statement on what is going to replace the Mac Pro - so we'll have to see how that goes. The Trashcan concept (which the Studio improves on considerably) as an "appliance" for running FCPx, Logic etc. could be a safer bet - and a more appropriate application of the Apple Silicon tech we've seen so far - than trying to fudge Apple Silicon into anything like the 2019 Mac Pro.
I suspect the Trashcan vs. Tower thing was an internal schism within Apple and that the 2017 iMac Pro was originally going to be the new Mac Pro. That surprise press conference in early 2017 was timed about right for the iMac Pro prototype to have just been unveiled under NDA to key players - and there was clearly blood on the carpet as a result.
The 2019 Mac Pro seemed to take expandability to almost ridiculous extremes (making the lower spec configurations disproportionately expensive), and could have been calculated to avoid direct competition with the iMac Pro (or, by 2019 when the iMP was abandonware, the high-end iMacs).
"Something else" would mean switching to Windows or Linux and potentially $$,$$$ worth of disruption & re-training if your workflow was based on MacOS-only software - that could easily dwarf the cost of equipment. If the management accept that you need a Mac and not a PC the issue goes away as long as it was costed in. If the management are in the habit of ignoring the consequential time & labour costs of saving money on the IT budget (i.e. business as usual) then, dude, you're getting a Dell (and probably lost that argument years ago).
My impression of the 2019 Mac Pro was, quite definitely, only worth considering if you're expensively committed to MacOS so the costs of changing outweigh the equipment cost. Otherwise, it's just a pretty Xeon W box with improved cable management. More powerful/specialised workstations are available at a price (more cores with scalable Xeon and AMD, specialist multi-GPU setups for CUDA/OpenCL workloads) as are far cheaper ones if you don't need such extreme RAM and PCIe capacity.
That's not a long-term solution for Apple, as it is gong to be a steadily shrinking pool as customers switch to PC one by one. Not to mention the growth of cloud computing and processing power on-demand (and Apple don't have a horse in the high-density-computing/server race).
Personally, though, I don't understand why any large-scale business is still considering the Mac Pro. Not because of the price, but because of the uncertainty - after 3-4 cycles of Apple introducing a radically different Mac Pro and then abandoning it (original cheesegrater - left to rot & discontinued in Europe for several years, 2013 Trashcan - never updated, 2017 iMac Pro - never updated, then dropped, 2019 Mac Pro - watch this space!)
Seriously, big businesses and institutions will often be working on tenders for future contracts that won't be awarded/signed for months and could last for years. Making a case for needing more expensive Macs shouldn't be a problem if you can specify & cost it, but how the heck do you do that when you have no clue whether there will even be a Mac Pro in 6 months time? Or you could spec a generic Windows/Linux workstation and be pretty confident that something comparable or incrementally better will be on the market for the next 3 years. One-man-band freelancers and enthusiastic hobbyists can turn on a dime, but if I were a more substantial enterprise with multiple employees to be trained and kitted out I'd have started phasing out Mac-only software once the Trashcan turned 4 with no updates.
Maybe for you. I use the NVMEs together in RAID for multi-stream ProRes 4444XQ. TB can be a bottleneck at times for some workflows.
hardly. This is a real use case scenario for me on the daily.
I'm a longtime Mac Pro user and care about looks and convenience. This is the problem with this level of generalization. Just cause it doesn't matter to you or someone else, doesn't mean it doesn't matter for other users.
At the end of the day, the value of a Mac Pro in its current iteration comes down to individual preference and need. Some value expandability via PCIE over others. Some have workflows that completely use hardware that others could care less about. There are folks have zero use for the Afterburner card or the AJA Kona video I/O, but they are integral in my use case and having them tucked away conveniently into a single tower provides more value than hooking up every single expansion component via TB and their own respective external enclosures.
I will always remember the crowds reaction after they announced the display stand's price. Still laughing out loud.And yet the most memorable part about this release was the $1000 display stand and the $400 wheels and the cost of the graphics card that was the entry price of the Mac Pro. $5900 for a W6900X?
It only matters in terms of uncertainty whether Apple is going to do another. (Which, this time, they have pledged to.)
As long as you know there will be an upgrade down the line, it doesn't matter much that they only update it every three years.
This is the core point that a lot can’t grasp because it’s so unlike GPU’s folks would buy in the past. Other companies are TRYING to get there… some of the newer AMD systems are marginally faster if you have an AMD GPU that talks to the AMD CPU, but it’s nothing compared to the performance of a system that doesn’t have to shuttle massive amounts of data to the GPU for rendering.
I don’t see such an upgrade happening. We haven’t seen that since the 1990s.
It doesn’t change that GPUs which are way faster at other workloads exist. Nor that, if the upcoming Mac Pro doesn’t support GPU expansion nor RAM expansion, it seems pointless to make. At that point, you’re talking about a Mac Studio, which already exists.
And why do they do it? To fit with the chairs? Or, perhaps, Mac is that bad at networking?
If apple is just a little bit pragmatic they will give us a MP that is the logical continuation of the mp2019. Same people running the show right? We’ll surely get a large expandable box, maybe the same enclosure even with mac, ultra and extreme options for the cpu and custom compute cards that accelerate Metal for e.g. raytracing in renderers. The compute cap. Must be above the 2019, so expect more than 100 TF combined. Maybe much more. T h, if this is not what the deliver, it is simply not a mac pro but rather a mac studio/thrashcan again. And obviously apple knows this. The question then becomes: can apple creat such a machine? Sure they can!
But co la it’s have switched, it’s short sighted to look at the cost of switching, use large businesses never look at the short term, they will look at the long term, retrain look at the long term as switch, then you have a better and more adaptable advantage to upgrade. The current Mac Pro negates your argument, because it’s easily upgradable, Apple wouldn’t make that for a handful of freelancers.
We await to see what the Apple silicon Mac Pro replacement is like, seeing as they have publicly committed to one.
you used "There’s no TB4 eGPU on Macs." as an argument which is irrelevant since no one stated that there is currently. not my problem if you got the argument wrong.
No need to continue talking about something that doesn’t and won’t exist. Save energy.
"for me" is anecdotal.
trashcan mac looks far better than cheesegrater. being able to easily unplug and share devices with other computers is far more convenient than internal addons.
tb4 LITERALLY exists today.
sorry but you're not really getting it. have a nice day.
You’re not getting that Apple uses UMA and eGPU doesn’t and won’t. Hence, no Apple Silicon drivers for eGPU.
Save energy for another subject.
many reasons. save money by moving the high performant machines to the edit bay with the appropriate setup that needs it the most. renovations to edit bay may cause down time so workstations are moved to temporary bays. IT may take the mac pro out for servicing so a temporary mac pro is moved in. list goes on.
Another what after a decade of flip-flopping between PCIe towers and external-upgrades-only trashcans? What three years when it's already late, and history has included 6 years between the Trashcan and the 2019 tower?
I think Apple are presuming that the "serious pro business" users can turn on a dime just like the consumers buying iDevices (...and we probably know more about the next iPhone than we do about the next Mac Pro).
Apple haven't even said, unambiguously, "yes, there will be a new Mac Pro with Apple Silicon", they've said something like "That leaves the Mac Pro, but that's for another day". Which probably means "a new ASi Mac Pro is coming" but is hardly a pledge - they're not gonna get sued if they basically rename the M2 Ultra Mac Studio as "Mac Pro".
I'll be honest with ya, if Apple hadn't made the 2019 Mac Pro, I very likely would've had to have left them. I run 2 w6800x Duos in it and 28 cores am still making new upgrade "about to install 32tb of ultra fast storage from OWC as well as another upgrade to my ram. The ONLY thing I don't like about it is that it's LIMITED to 28 c ores and it's very slow CPU architecture when it comes to my workflow...but as I said before, it's a realtime 3D artist's dream which is the only thing I use it for, 3D animation, VFX and Rendering.
Apple doesn’t generally tease things they don’t deliver on. It’s possible the Mac Pro has since been canceled in favor of a higher-end Studio, but I don’t believe that was Ternus’s plan when he said that.