MP 1,1-5,1 What machine do you expect to be using in 2025?

[tsialex]

You are assuming that the Apple SoC used in the upcoming Mac Pro will be similar to the M1 or M2?

Pro users won't go for that, we want PCI-E slots and upgradable memory, Apple learned it's mistakes with the 2013 Mac Pro. Tho the Mac Studio maybe Apple's way of testing the waters on never releasing the Apple SoC Mac Pro.

Pro towers like the 2019 Mac Pro don't sell in large numbers, Apple may just cut us loose.

Mac Pro as we know was always financed by the volume of workstations and servers sold by Intel, Apple always iterated and tweaked on the reference designs provided and already payed for by Intel selling tens of millions of server CPUs. Now Apple have to finance (as in ROI) everything itself.

Apple investment with Apple Silicon to this day is completely focused on that Apple can sell high volume, as in millions of units - like iOS devices, MacBook Airs, MacBook Pros and some devices that have less volume but are really needed to make a complete line of products and not to difficult to implement with the current chip designs, like the 24" iMac, Mac mini (exactly the same M1 CPU of Air/13" MBP) and Mac Studio (same or doubling the M1 Max CPU design of 14" and 16" MBPs). Apple don't even have a 27" iMac substitute to this day, almost two years down the road of Apple Silicon.

If you are taking clues from the current silicon designs and how Apple is implementing the support software, you can easily see for yourself that RAM and CPU upgrades are not even possible with the M1 and M2 architecture, unless is something along the lines of a much slower level 2 RAM buffer or or a full CPU on a card/module that is removable, that you will have replace the full card/module when upgrading - something that I really don't think that Apple will do, but it's possible.

Apple will have to change radically the M2 design to reimplement the 2019 Mac Pro with Apple Silicon, which I really doubt it since even if you account the number of Mac Pro ever sold to this day wouldn't have a return of investment necessary. The most probable is that we will get something like a doubled Mac Studio Ultra on steroids, with new I/O and slots for additional network/audio/storage cards but not upgradeable RAM or CPU.

 

[mode11]

OP, you need to add an option for: Brand New Mac Pro 7,1

I could do, but when creating this poll I was interested in peoples' thoughts on the upgrade path from the 5,1 Mac Pro. My assumption was that anyone still using one of these as their main machine, who hadn't yet pulled the trigger on a new 7,1, was unlikely to do so at this point. The AS MP might be announced in a matter of months; this will a) show Apple's hand in terms of where the MP is going (including its price), and b) eventually lead to 7,1's becoming available at affordable prices on the second hand market.

I don't expect the price of 7,1's to collapse - they'll continue to be useful production machines for some time to come. Also, the AS MP is likely to be both expensive, and disappointing in its upgrade potential. Obviously user-replaceable CPUs and RAM DIMMs are a thing of the past. It's also hard to see Apple making a particularly custom SoC for the new MP, as this would be a huge investment for a product line they've repeatedly left to stagnate, and almost certainly hoped to kill off (with the iMP, also left to stagnate). MP profitability likely only works for Apple if they keep the machine untouched for 4-5 years after release, to spread the development costs.

On the other hand, perhaps the 7,1 does give us a guide here. Apple essentially said "OK, we don't particularly want to make the Mac Pro, but if we have to in order to keep our highest-end users, then we'll go all-out and charge accordingly". Just giving us a machine with 2x Ultra's in it seems unimaginative. That approach also means those who primarily want GPU power have to pay for an absurd amount of CPU power to get it. I also doubt that Apple want to get into a GPU arms race with Nvidia and AMD, companies that specialise in that area, and unlike Intel, haven't spent years in the doldrums. It also doesn't solve the problem the 7,1 was designed for - to internally house PCIe cards for audio, video, RAID, networking and so on. For those reasons, the AS MP would surely need to be able to support PCIe, if it's to have any relevance vs. a Mac Studio Ultra.

 

[DearthnVader]

Mac Pro as we know was always financed by the volume of workstations and servers sold by Intel, Apple always iterated and tweaked on the reference designs provided and already payed for by Intel selling tens of millions of server CPUs. Now Apple have to finance (as in ROI) everything itself.

Apple investment with Apple Silicon to this day is completely focused on that Apple can sell high volume, as in millions of units - like iOS devices, MacBook Airs, MacBook Pros and some devices that have less volume but are really needed to make a complete line of products and not to difficult to implement with the current chip designs, like the 24" iMac, Mac mini (exactly the same M1 CPU of Air/13" MBP) and Mac Studio (same or doubling the M1 Max CPU design of 14" and 16" MBPs). Apple don't even have a 27" iMac substitute to this day, almost two years down the road of Apple Silicon.

If you are taking clues from the current silicon designs and how Apple is implementing the support software, you can easily see for yourself that RAM and CPU upgrades are not even possible with the M1 and M2 architecture, unless is something along the lines of a much slower level 2 RAM buffer or or a full CPU on a card/module that is removable, that you will have replace the full card/module when upgrading - something that I really don't think that Apple will do, but it's possible.

Apple will have to change radically the M2 design to reimplement the 2019 Mac Pro with Apple Silicon, which I really doubt it since even if you account the number of Mac Pro ever sold to this day wouldn't have a return of investment necessary. The most probable is that we will get something like a doubled Mac Studio Ultra on steroids, with new I/O and slots for additional network/audio/storage cards but not upgradeable RAM or CPU.

You know a lot about what you are saying, and I here you, it's just that non-upgradable RAM is a deal breaker for those of us that need it. Apple set the bar pretty high with the 2019 MP( 1.5TB ).

They can't afford to put that much RAM on the Die, and it would be huge and power consuming and god knows what the TPD would be. They couldn't sell enough of the chips to make money on the production of them.

Either the Apple SoC Mac Pro will have upgradable RAM or is vaporware and they will never produce it.

I'm more inclined to think it will be as you say, just a Studio Ultra, but I'm hoping......

 

[mode11]

Tho the Mac Studio maybe Apple's way of testing the waters on never releasing the Apple SoC Mac Pro.

Pro towers like the 2019 Mac Pro don't sell in large numbers, Apple may just cut us loose.

Ordinarily, I would assume this. But Apple have already announced the future existence of the Mac Pro, and though I expect the Mac Studio has sold quite well, Apple won't be certain how many people are still holding out for the MP. Bear in mind that Apple discontinued the 27" iMac at the same time, so some of those sales will have been picked up by the Studio (or M1 mini), paired with either Apple's 27" 5K display or a mainstream 4K monitor.

In my opinion, the iMac Pro was clearly intended to replace the 6,1, and was well into development before Apple decided to do a 180. Something convinced Apple in early 2017 that if no proper workstation (i.e. one able to house multiple high-powered GPUs) were available, pros would reluctantly give up on macOS and start moving to Windows. Apple's roadmap secrecy was untenable for businesses that needed to know whether the company had any long-term commitment to serving the workstation market.

Apple are obviously loath to a) admit they get anything wrong or b) pre-announce a product. So in doing both, we can assume they felt it unavoidable. They knew it would take 2-3 years to design / set up production for the new machine, starting from scratch. They also knew that if the 6,1 just carried on untouched for another few years, with Apple silent all that time, that there would be no high-end Mac market left to buy the 7,1 when it came out.

Having recently gone through that process, it seems strange for Apple to just dump the tower again. You'd think that it must be in the late stages of development, if the original intention was to reveal it in late 2022. Though I suppose if MS Ultras are selling like hot-cakes, they could cut their losses on development and quietly shelve it.

 

[DearthnVader]

Ordinarily, I would assume this. But Apple have already announced the future existence of the Mac Pro, and though I expect the Mac Studio has sold quite well, they won't be in a position to know how many people are still holding out for the MP. Bear in mind that Apple discontinued the 27" iMac at the same time, so a few of those sales will have been picked up by the Studio (or an M1 mini), paired with either Apple's 27" 5K display or just a mainstream 4K monitor.

In my opinion, the iMac Pro was clearly intended to replace the 6,1, and was well into development before Apple decided to do a 180. Something convinced Apple in early 2017 that if no proper workstation (i.e. one able to house multiple high-powered GPUs) were available, pros would reluctantly give up on macOS and start moving to Windows. Apple's roadmap secrecy was untenable for businesses that needed to know whether the company had any long-term commitment to serving the workstation market.

Apple are obviously loath to a) admit they get anything wrong or b) pre-announce a product. So in doing so, we can assume they felt it was unavoidable. They knew it would take 2-3 years to design / set up production for the new machine, starting from scratch. They also knew that if the 6,1 just carried on untouched for another few years, with Apple silent all that time, that there would be no high-end Mac market left to buy the 7,1 when it came out.

Having recently gone through that process, it seems strange for Apple to just dump the tower again. You'd think that it must be in the late stages of development, if the original intention was to reveal it in late 2022. Though I suppose if MS Ultras are selling like hot-cakes, they may cut their losses on development and quietly shelve it.

It takes like 6 months for the PC would to develop a product based on x86_64. It takes Apple 12 months to design Apple SoC's for the iPhone, but then the Pro market( Workstations ) will never do the pure volume that the iPhone does.

We are the bastard children when it comes to Apple products, we just end up with the hand-me-downs.🤣

 

[mode11]

We are the bastard children when it comes to Apple products, we just end up with the hand-me-downs.

I agree. I think the move to Apple Silicon is primarily about porting macOS to the iPhone's hardware / software stack. Whatever the energy efficiency benefits (which are substantial at present), the cost savings are key. Though given Apple's industry-leading CPUs, it's hard to blame them. The only issue is that whereas developing an SoC suitable for a laptop or iMac can be done with a beefed-up iPad SoC, the Mac Pro - the nichest of Apple's products - would require something substantially bespoke. The obvious concern is that Apple will chose to simply discontinue the line instead.

Though again, why make the 7,1 in that case? They could have just relied on the iMac Pro to hold people over until the Mac Studio (or even updated the 6,1 with newer Xeons / Radeons / TB3 - how hard would this have been?).

 

[bmoraski]

Ill be using my 5,1 ( Mojave ) until it completely dies. I have it setup to do everything i need it to do. All my photo's, home movies ( converted all old VHS and Hi8 videos to digital ) , all my music ( CD's etc ) and Garage band for guitar tinkering. Such a beautiful machine and i always wanted the cheese grater tower. Have had a little over 2 years. Im also a collector. I still have my original 17" imac corduo upgraded to 2 gig of ram. lol

 

[mattspace]

Though again, why make the 7,1 in that case? They could have just relied on the iMac Pro to hold people over until the Mac Studio (or even updated the 6,1 with newer Xeons / Radeons / TB3 - how hard would this have been?).

Because the 6,1 was a failed paradigm, not a flawed implementation of a good paradigm. It was a failure when it was brand new, it was a failure when it was 1 year old, it was a failure when it was still the flagship mac 5 years later.

The 6,1 made Apple a laughing stock in professional circles. Bruised ego & butthurt gave us the 7,1.

 

[tsialex]

Because the 6,1 was a failed paradigm, not a flawed implementation of a good paradigm. It was a failure when it was brand new, it was a failure when it was 1 year old, it was a failure when it was still the flagship mac 5 years later.

The 6,1 made Apple a laughing stock in professional circles. Bruised ego & butthurt gave us the 7,1.

I don't see the late-2013 Mac Pro as a failed Mac and for some workflows, it's just perfect. As a MacPro5,1 successor, yes, it's a complete failure. A little less ego from Apple, better marketing and focus on what the platform did best would have made the MacPro6,1 a success as something else on the Mac line.

Looking back, it was the completely wrong successor for the MacPro5,1, but it's a great and iconic Mac. If you look from the perspective as something as a Mac Studio prototype ahead of time, it's a silent numbers cruncher and a lot of people really love it.

 

[mode11]

As a fan of the G4 Cube, I've always seen the 6,1 as a beautiful piece of design, if not something I'd actually buy (unless / until it becomes cheap, purely as a collector's item).

When I said 'why make the 7,1', I wasn't implying they shouldn't have - it was the correct decision (as I've written about above). I meant that if the ultimate plan was for the Apple Silicon-based Mac Pro 8,1 to be something like a beefed-up Studio, why introduce their most expandable machine ever in between the 6,1 and the 8,1? It set expectations for future machines, especially after making a big deal about the time they'd spent researching the workflows of high-end users. To go back on this will create a lot of unnecessary bad will. Their thinking seems all over the place on this.

 

[mattspace]

I don't see it as a failed Mac and for some workflows, it's just perfect. Looking back, it was the completely wrong successor for the MacPro5,1, but it's a great and iconic Mac. If you look back at the late-2013 Mac Pro not as a MacPro5,1 successor, but something as a Mac Studio prototype, it's not really a failure and a lot of people really love it.

Yeah, but you could just as easily call the Mac Mini a "Mac Studio Prototype", cause the Mac Studio is just a more powerful Mini - it's closer to that, than it is to the paradigm of the 6,1 (dual GPU with only one wired for display etc).

The 6,1 brought nothing to the table that was necessitated by its particular design choices - none of the compromises it brought came with a benefit that required those compromises, and those compromises included endemic thermal failures - just like the G4 cube had endemic thermal failures.

It was a bad design - a pretty bit of three dimensional decoration, if you're into smeared fingerprints, but as a design exercise, it fails on every criteria it's intended to address.

 

[mode11]

The 6,1 was quiet and compact, yet fairly potent at the time. Good for those wanting a fancy SFF computer, though admittedly irrelevant to production use.

The appeal to Apple, apart from showing the world that post-Steve Apple could still 'innovate' (yes, yes, I know, but it was certainly striking), was probably that the Mac Pro would take up much less space in Apple Store storerooms, and be cheaper to make and transport.

The 7,1 was possibly an over-correction to the 6,1 (though I'm sure some will have all slots filled). Many people would be happy with 4 slots, less exotic construction, and lower cost. But I guess a sensible tower PC just doesn't have the design kudos Apple craves.

 

[tsialex]

The 6,1 was quiet and compact, yet fairly potent at the time. Good for those wanting a fancy SFF computer, though admittedly irrelevant to production use.

To this day there are co-location services installing brand new late-2013 Mac Pros in datacenters, very recently I saw a picture of several pallets of brand new late-2013 Mac Pros arriving to a datacenter and someone commented that around 600 units were acquired from Apple. For some use cases like virtualization, CI and similars, the MacPro6,1 is just perfect. Hey, GitHub uses troves of MacPro6,1s for CI.

Obviously this will change with Ventura and new the Xcode releases that won't run with MacPro6,1, maybe that why Apple is offering cloud CI now, but calling it irrelevant to production is just not true.

 

[mode11]

To this day there are co-location services installing brand new late-2013 Mac Pros in datacenters, very recently I saw a picture of several pallets of brand new late-2013 Mac Pros arriving to a datacenter and someone commented that around 600 units were acquired from Apple. For some use cases like virtualization, CI and similars, the MacPro6,1 is just perfect. Hey, GitHub uses troves of MacPro6,1s for CI.

Obviously this will change with Ventura and new the Xcode releases that won't run with MacPro6,1, maybe that why Apple is offering cloud CI now, but calling it irrelevant to production is just not true.

What makes a 6,1 superior to a 2018 Mac mini, for this purpose? More cores? GPU power?

 

[tsialex]

What makes a 6,1 superior to a 2018 Mac mini, for this purpose? More cores? GPU power?

Xeon processor with loads of L3 cache, 12-cores, 128GB of ECC RAM, replaceable/upgradeable storage.

Since the GPUs are almost not used with Virtualization/CI workloads, the thermals are also much better than the high end 2018 minis when rack mounted.

 

[SpotOnT]

I agree. I think the move to Apple Silicon is primarily about porting macOS to the iPhone's hardware / software stack. Whatever the energy efficiency benefits (which are substantial at present), the cost savings are key. Though given Apple's industry-leading CPUs, it's hard to blame them. The only issue is that whereas developing an SoC suitable for a laptop or iMac can be done with a beefed-up iPad SoC, the Mac Pro - the nichest of Apple's products - would require something substantially bespoke. The obvious concern is that Apple will chose to simply discontinue the line instead.

Though again, why make the 7,1 in that case? They could have just relied on the iMac Pro to hold people over until the Mac Studio (or even updated the 6,1 with newer Xeons / Radeons / TB3 - how hard would this have been?).

As far as I can tell, the whole concept of personal ownership is declining. I wouldn't be surprised if hardware (whether we are talking computers, phones, or cars) become a service industry, where you pay a subscription fee to have the "capability" and the company will just supply you with the hardware.

No reason for Apple to invest in a Mac Pro tower in that world. Everything will just be in sealed boxes, with hardware optimization for the popular codecs of the day.

 

[Kimmo]

Ill be using my 5,1 ( Mojave ) until it completely dies.

Speaking of Mojave, Apple could do their customers (and themselves) a favor by issuing a patch for the recent security breach:

CIS Advisories

CIS Advisories.

www.cisecurity.org

If you're a 5,1 owner running Mojave who is poised to make a significant Apple purchase in the near future (the new thing, or a 7,1) the issuance of a Mojave patch would go a long way to showing that Apple gets it.

Like Steve Jobs before him, Tim Cook maintains a public email address. If you're a Mojave user, or just someone who's concerned about the security and privacy of Apple's operating systems, a note to Mr. Cook requesting a Mojave patch could help. His email is tcook@apple.com.

 

[WayneStewart]

I just replaced my 2013 MP with a Mac Studio. I’ll keep my 2012 Mac Pro around for running some older versions of Adobe software.

I’m feeling the urge to replace my 2019 Macbook Pro but I’ll fight that until at least the M2 comes out.

 

[deconstruct60]

Apple will have to change radically the M2 design to reimplement the 2019 Mac Pro with Apple Silicon, which I really doubt it since even if you account the number of Mac Pro ever sold to this day wouldn't have a return of investment necessary. The most probable is that we will get something like a doubled Mac Studio Ultra on steroids, with new I/O and slots for additional network/audio/storage cards but not upgradeable RAM or CPU.

Apple doesn't have to radically change. Apple needs to do some work. 2024-5 is 2-3 years from now. Even if M2 doesn't add it at gen 2 by gen 3 they may be a few more options.

If Apple picks up PCI-e v5 (or v6 ) in 2024-25 then that would open door to also implement CXL. Won't be the same as first line, RAM but something like

https://www.servethehome.com/sk-hynix-cxl-2-0-memory-expansion-modules-launched-with-96gb-of-ddr5/

would allow a card that used PCI physical layer to work with a potential driver for macOS to provide specailizesd RAM that was faster than storage drive options. ( e.g., faster page RAM for file system. Or an app specific RAM SSD or .. .)

UltraFusion connector is fast enough to add an I/O specific die that doesn't have to be in the mac laptops and Minis (or Studios ). Or Apple can just switch some of the die space used on Thunderbolt over to PCI-e v4/5 some of the > 1 die packages UltraFusion is a demonstration that Apple is willing to spend limited money on supplemental chips for a larger, multiple chip M-series package.

So far Apple isn't a big fan of CXL , but if they changed their mind now then 3-4 years down the road something could come out of the design/development pipeline. It wouldn't be dirt cheap expandable RAM but they are provision contemporary x8 - x16 PCI-e slots at that far in the future there could be an option. [ A couple of years ago CXL had rivals, but the competitors are dying off. Most architecture implementers are on board at this point. 2-3 years from now it should be closer to be more commonly implement. If Apple continued to swim against the tide of the PCI-e progression and CXL they would be on an increasingly isolated island even if the network/storage space.

Samsung Memory-Semantic CXL SSD at FMS 2022 Powered by AMD-Xilinx

At FMS 2022, the Samsung Memory Semantic SSD uses AMD Xilinx and a PCIe Gen4 NVMe SSD to build a CXL SSD running CXL.io and CXL.mem

www.servethehome.com

]

3-4 years down the road, CXL 2.0 is going to be commonplace in the high end workstation offerings that the Mac Pro will have to compete with. Unless there is some huge mismatch with Apple's memory implementation semantics they'd be on thin ice if they don't have it.

Even if the primary CPU cores , GPU , and RAM are locked down there are evolving standards that let accelerators on PCI-e PHY layer participate. Apple blocking that buy that a ton of upside if gong to allow PCI-e PHY slots anyway. The "extra power' has been spent just by provisioning the slots.

That is also separate from letting 3rd parties back into the graphics driver stack/subsystem.

 

[ssj92]

For me I'll always be using a "Windows PC" but I also use Macs so I voted for AS Mac Pro for now.

I am waiting for them to unveil the AS Mac Pro to decide whether I will go Mac Studio or AS Mac Pro.

I would love to get a 7,1 Mac Pro but at this point I need something that can do 8K HVEC and even my iMac Pro struggles (Even 7,1 MP struggles as seen in some threads here).

If next gen AMD GPUs are fully supported by macOS and work in 7,1 Mac Pro then maybe I'll go that route, but at the same time I could get a eGPU enclosure for my iMac Pro and use those GPUs there and get an AS Mac Pro or Studio so we'll see.

I really like the cMP because of its expandability and the fact it has survived this long running Monterey. I even had put a RX6800 in mine which was incredible. I know I'd get the same enjoyment from a 7,1 Mac Pro as I did from cMP but prices need go come down more then perhaps I'll get one to play with. Maybe in 2025 I'll get a 7,1 to play with. 😁

 

[macguru9999]

Speaking of Mojave, Apple could do their customers (and themselves) a favor by issuing a patch for the recent security breach:

CIS Advisories

CIS Advisories.

www.cisecurity.org

If you're a 5,1 owner running Mojave who is poised to make a significant Apple purchase in the near future (the new thing, or a 7,1) the issuance of a Mojave patch would go a long way to showing that Apple gets it.

Like Steve Jobs before him, Tim Cook maintains a public email address. If you're a Mojave user, or just someone who's concerned about the security and privacy of Apple's operating systems, a note to Mr. Cook requesting a Mojave patch could help. His email is tcook@apple.com.

or you could just upgrade to dosdude catalina .....

 

[macguru9999]

Its amazing that we are talking about this now because most of us have been happy with our 5,1 macs, 10-12 years after they were manufactured. I hope the 7,1 mac pro can last that long ....

 

[fatespawn]

I assume I'll still be running my 4,1>5,1 MacPro. I had a single processor version and 2 years ago found someone selling a dual processor version for a couple hundred bucks on Facebook. Couldn't pass it up. Now I have two of these fine machines. New processors, new thermal paste, upgrade, upgrade, upgrade. They still run great. I'm not in the "production" business, so I assume my next computer will be my son's hand-me-down Ryzen homebuilt from the basement. Considering I'm still using a 13 year old computer for my daily-driver, it doesn't matter much.

 

[scottrngr]

I'm waiting for whatever comes out in 2026! 🤣

 

[krakman]

I think I will still be using my 2009 4.1 >5.1 Mac Pro.

It's fast enough for email, browsing, editing photos.

I will use it for as long as it powers up.

However I will have to buy an apple silicon Laptop at some point soon almost of my work is at client's venues.