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.
MP 7,1 Mac Pro 8.1 - WHAT IF...
- Thread starter maikerukun
- Start date
-
- Tags
- 7.1 mac pro next generation
- Sort by reaction score
Are you saying that if someone plugged in e.g. an RX580 eGPU to an M1 Mac via Thunderbolt right now, they would have accelerated display output from the eGPU?
In another thread someone posted the exact quote from what Gruber said.
There is huge leap from "will look identical" ( which is extremely suggestive is simply talking about just the external appearance. ) to logical board looks exactly the same , all the internals look exactly the same. The connectors are all the same.
Which if go to the next part which says there is no user-upgradable RAM, that can't be true. The DIMMs slots are gone. So that does not look the same.
If the DIMMs connectors look the same but are gone, then the same thing can be true for the MPX connector.
The most can inference from just the external case is that the slot layout spacing is the same. It actually doesn't have to have the same number of slots on the motherboard. Where there is no slot could be just a vestigial slot hole present (with nothing really behind it).
There is an obtuse reference to Graphics but doesn't necessitate a completely backward compatible MPX module.
Not sure where the reference to the Max comes from. If beliving Gruber then he has stated that the Mac Pro starts at Ultra.
Even if Apple was going to build some new proprietary boot on AS MPX modules .... they would boot in a non UEFI context. That wouldn't mean they would automatically be compatible with the older Mac Pro. Nor does that mean those drivers got automatically back ported to the macOS intel side of the driver implementations. ( it sure has not worked in the other direction automagically).
If Apple is having to sponsor and/or spend money making a reference card work with an special ROM are they going to really spend even more money so it is a UEFI/iBoot crosswitching card?
If the new Mac Pro has the Thunderbolt controllers inside of the SoC then MPX is a largely a solution in search of a problem. MPX solves the problem of how to get inputs to the TB controllers mounted near the outer edges of the system. If the TB controllers are in the SoC that really is closer to the opposite placement for the controllers.
The AS Mac Pro will have an iGPU. The Intel Mac Pro doesn't have an iGPU. The electrical display ouput path layout would be substantively different. The display sources are located in completely different locations of the system
The AS Mac Pro would not need a complicated DisplayPort re-route mix and match networking to match discrete TB controllers to various different GPU source locations. It would all flow out of the SoC in point to point connection to the external physical sockets. A significant amount of wire routing and switching could be tossed to save money.
apple could spend extra money to remove the TB controllers from their SoC. Go back to spending money to buying Intel discrete TB controllers (when they have their own good implementation) and continue to keep spending more extra money on a more complicated motherboard driving the system cost higher.
Or could just go to a system where each GPU ( iGPU and dGPU) just drive video out through their own associated output ports. ( the dGPU TB controllers would need some PCI-e input but there is no display out to reroute through the internals of the Mac Pro). If it is a x16 PCI-e v4 connect the card could just directly bleed the TB output off the that input to the card. Don't really need that part of the MPX connect either. But if want ot pretend it have an impoverished x16 PCIe- v3 feed just like the Mac Pro 2019 ... that is an option.
I really hope the entire motherboard itself is replaceable.
Given the prices of the wheels and the stand for the Pro Display XDR, versus the presumably relatively cheap price of the Ultra-class SoC, you can assume that the 8.1 case itself is potentially more expensive than the computer inside it.
I hope that going from M2 Ultra to an eventual M3 Ultra will not require a lot of people to purchase their 3rd identical giant (beautiful) lump of machined aluminium.
Given the prices of the wheels and the stand for the Pro Display XDR, versus the presumably relatively cheap price of the Ultra-class SoC, you can assume that the 8.1 case itself is potentially more expensive than the computer inside it.
I hope that going from M2 Ultra to an eventual M3 Ultra will not require a lot of people to purchase their 3rd identical giant (beautiful) lump of machined aluminium.
I have not seen a proper leak from Gruber in years. Has he released something about the Mac Pro? I haven't seen it.
I'm referencing what Gurman wrote in his Power On newsletter that was mentioned in this thread.
It would sure be unprecedented if Apple tried to keep the internals the same on the new Mac Pro. Empty Xeon sockets that you couldn't use and RAM slots that just took up space. No one is expecting this. I'll go as far as not even including it as a "black horse" in the race.
Correct. It does not. But as opposed to my last 3 or 4 posts or so, in my last post that you quote, as I state in the beginning, I'm starting with Gurmans update and looks at things from a positive perspective. It's all speculation.
It comes from the Mac Studio Max version. That computer holds its own for many users and should Apple release a Max version, it would be even cheaper than the Ultra and allow even more people to opt for the Mac Pro.
Just to be clear: my latest post assumes MPX modules. In previous posts I have leaned more towards a specialised Apple Silicon accelerator card that I wouldn't even classify as a GPU. But now we are talking about "graphics cards" and a similar looking AS Mac Pro, so I'm assuming MPX for the argument.
Yes, Apple could release new MPX modules and exclude the only other computer on the market that, that they themselves make, and exclude those customers. As you say, perhaps with a driver trick: "if CPU = Intel, make incompatible //#AppleSiliconFTW"
All in all, interesting feedback and only time will tell. It's all speculation. First prudent conservatism, and now cautious optimism.
I would not put any amount of money on there being an entry level AS Mac Pro that costs less than the highest spec Mac Studio.
Pros didn't save Apple.
I feel like this has to be harped on every time someone brings it up. You know what saved Apple from bankruptcy? Axing half their products and then shipping the iMac, a consumer product. Know what saved them after that? The iPod. After that? The iPhone.
Desktop publishing was a major segment of Apple's portfolio in the late 80s and early 90s, but they were going to go out of business in the 90s. It wasn't sustainable. They saved themselves by producing something that was, if aimed at any market segment, built for education. But mostly just chased new users.
Doesn't mean that professional and prosumer use cases aren't a very good business that Apple should remain engaged in, but the history of professionals keeping Apple alive during the dark times just is incorrect, and it suggests the idea of the company "owing" people for their business, which isn't how it works either. Either Apple products work for you, or they don't. It's a company, it doesn't have feelings and it doesn't owe consumers anything.
While I get your point, the restaurant analogy is a terrible one to use here. Food service is an incredibly low margin business. That $300 steak you get might seem like they're ripping you off, but they are likely losing money on that versus making it up by selling a bunch of cheaper steak frites, because they can't mark up the $300 steak enough to make money off it (the price would be unacceptably high.) Alcohol is the easiest way for restaurants to make up their margins. They're not ripping people off, it's the customers who would refuse to pay the real cost of their meals outside of alcohol.
To the same degree, besides the fact that Apple is a company that is beholden to shareholders to make ever-larger profits, which is a fools errand long-term, the simple bill of materials for computers usually doesn't factor in the cost it took to develop a million prototypes, or marketing, or the support staff that will be needed, or the training. Upgrades are a significant way for them to make that back—you are paying for the innovation and design by getting charged an extra 50% or whatever for RAM than it'd cost you to get it yourself.
I assume the M2 will have something closer to a 192-256GB RAM ceiling, and that's certainly most for even most pros, but obviously there will be the 1TB+ crew that will get left out in the cold (scientific modeling is a big example—the only way to get performant models is to have the entire thing in RAM.)
Until the day-of rumors, I'm not going to believe Apple will keep third-party GPUs around on Apple Silicon. Doesn't make sense to get the hopes up. But it's also a really weird time for Apple where they're clearly being forced to stagger their products by months or years longer than they wanted, and if the Mac Pro is a end-of-2023 product like its predecessors I can see the requisite stuff showing up in the MacOS 14 betas at WWDC. Doesn't mean I'd love to be running an early MacOS version on those new machines until the kinks got worked out with a few point updates, of course.
Gruber doesn't have to generate clicks with giant rumors drops and benefits more from a friendlier relationship with Apple. The most he does is mention when his sources agree with a rumor, which is how he prevents them from getting burned like Gurman (and really Gurman himself and the other Apple leakers seem to have lost a lot of their access. Presumably Apple's anti-leaking efforts have been successful. Some of their "totally missed it" predictions seemed very calibrated in a sort of "leak the false info so we see who squawked" kind of way.)
In short, don't trust the rumors. Especially since they could be months out of date and lacking all context by the time they make their way to the light.
So you don't think there will be any AS Mac Pro below $7999? Or do you mean below base spec Ultra?
And if so, what is your reasoning for this? What do you feel is Apple's reasoning?
I see no reason why the AS Mac Pro shouldn't start with Apple Silicon Max. It's more than enough for many use cases that would still like the expandability and/or tidiness of a Mac Pro.
No - AFAIK the topic is what could a Metal card do with no display output. Either case would still require drivers though.
Basically if you plugged an RX580 into a Mac that had drivers (e.g. an Intel Mac) you could use the GPU _without_ having it plugged into a display.
Bzzzt. That isn’t what saved apple. Provably false statement. Cutting products doesn’t bring income. Simple as that. Selling products brings in income.
Who did apple market to to save it and buy products with? Think different die hards. The ad campaign was as much a plea as it was a love letter to the people that think different.
Yes, making a good set of products is important, but more important is actually selling them to bring income. And you have to sell to someone…the only someone’s left when all others ran away were the think different die hards.
He said the iMac and it's pretty well documented that it was the iMac G3 that saved Apple. The power of the Think Different campaign was that it appealed across user segments, the only common thread was that it spoke to people who selected their computer themselves rather than having it imposed on them by corporate policy. Which is a way of saying that everyone with any level of creativity, whether actual or aspirational, felt it was talking to them. But the intent of challenging you to "think different" was to put you in a state of mind where you bought an iMac rather than a grey slotbox. And it worked.
Youre missing the point. They leaned on their 'think different' demographic. It's not just Mac pros. But that think different is highly influenced with their rabid loyalty with a HUGE enthusiast component (many of whom happen to be pros). That is the crowd that cheers when they see a Mac laptop in independence day. And that laptop was in Independence Day because of the influence of the Mac Pro (back then power Mac) halo set of enthusiasts and pros.
No Pros, that goes away.
Cutting products did stop losses which was a necessary prerequisite to regain profitability. Not only was the product line bloated, it was confusing as hell. It's so much easier today to point people to which mac might work for them because each mac has its intended audience/use case. It's mind boggling how many different models of Performa's there were. This is to say nothing of their other hardware and software efforts at the time.
Now the whole thing about the Think Different campaign is that it was a love letter to creative individuals as such, not to any particular computing use case. In fact, the campaign itself didn't talk about the products (since this was pre iMac). The iMac really did get the ball rolling once it was released. The release of the B&W G3 PowerMacs, OS X, PowerBooks, Final Cut, G4's, G5's, Logic, Aperture helped to build a solid foundation for creative pros, but it was such things as iLife, iMacs, iBooks, iPods, and iTunes that brought Apple back to the consumer markets in a big way. Both the pro and consumer markets are vital to Apple and Apple could not have survived without both.
Now whether the Apple of today thinks it still needs the high-end creative pro is a different story. I am hoping that the Creative Pro workflows team is working out the strategy for the Mac Pro and is getting serious about regaining lost ground, but only time will tell.
Ahem...
Now back to the dreaming...
I think Apple should issue a M#WS CPU chip (WS for workstation) alongside a G# MPX PCIe card for GPU. All SOC functionality not related to CPU/GPU would be included on a T3 controller SOC including I/O for expandable memory. Basic 32 core ML would be built into the T3, but there would be the option of an expandable ML MPX module for those who need it. Basic media encode/decode would be present on the T3 but could be expanded via an Afterburner 2 card.
There would be four processor slots for the M#WS CPU. Each CPU would have 16H/4L so that the total core count could range from 20-80 CPU cores. The PCI configuration would be similar to the 7,1 just updated to the latest generation.
The memory system would work similar to the 7,1 but would feature 8 memory channels instead of six.
A minimum spec machine would be:
1x 20 core M2WS CPU (16h/4e with 3 empty CPU sockets)
T3 controller with 32c Neural Engine for ML and M2 Max equivalent media encode/decode engines
64GB RAM
1x 64 core G2 MPX with 32GB VRAM with raytracing 2xHDMI 2.1 and 2x Thunderbolt
1TB storage with 4 M2 slots
6 built in Thunderbolt ports (2 on top, two in back, plus the two on the GPU)
4 USB A ports
SPDIF optical i/o
2x 100Gb Ethernet built in
Mic/line input
High impedance headphone jack
$5500 Tower, $6000 rackmount
Options
up to 4 CPUs for a total of 80 cores (64h/16e) (1x, 2x, and 4x configs supported)
up to 2TB RAM
up to 64TB internal storage
Neural Engine PCI card with 128 cores
G2x PCIe GPU with 128c 64GB VRAM with raytracing support
G2Pro PCIe GPU with 256c 128GB VRAM with raytracing
Afterburner 2 card with 8 video decode and 16 video encode engines and 8/16 ProRes decode/encode engines
Thunderbolt/USB4 I/O card with 4 USBC ports
Legacy I/O port card with 2x FW800, 2x FW400, 2x Thunderbolt 2
So a max spec pro could have:
4x M2WS CPUs for 80c (64h/16e)
2 TB RAM
64 TB storage
Afterburner 2 with 8/16 video decode/encode and 8/16 ProRes decode/encode
2x G2Pro PCIe GPU for 512c with 256GB VRAM
Thunderbolt I/O Card
$Arm+Leg and/or kidney
An even lower price minimum config could have a Mac Studio based motherboard design and would be sold alongside the full Mac Pro.
M# Ultra SOC
64, 128, 192 GB RAM options
1-64TB storage in 4 ssd slots
6 PCIe expansion slots
4 thunderbolt ports
1 HDMI port
10Gb ethernet port
2 USB A ports
mic/line input
High impedance headphone jack
$3750
The pricing is kind of impossible to decide on if you think there's any possibility to the rumor that the Mac Studio is going away (and also missing is the question of what the replacement of the Mac mini Intel looks like.)
I think a lower starting price is possible even with the Mac Studio, but I wouldn't expect it to be any lower than $5K, and like the 7,1 you'd have to put $1K into it to get a reasonable configuration for most people (bumping the storage, GPU, and RAM costs you $800 right there, and back when it shipped with a 256GB SSD and the comparatively much worse 580X you'd probably have wanted even more, even if you did things like RAM aftermarket.)
Think Different was entirely about rebranding Apple inside and out, not trying to directly sell Macs (hence why it featured no products). From Jobs himself: "We have to prove that Apple is still alive [...] and that it still stands for something special" (to TBWA\Chiat\Day ad man Lee Clow.) "It was directed not only at potential customers, but also at Apple's own employees." (Isaacson, 328). It was an expression of Jobs' personal philosophy, not trying to speak directly to professional Mac users.
The idea of the Mac Pro as a Halo car has been brought up a lot, but if that were true I'd argue Apple would clearly be in much worse shape now than it is, given that there hasn't been a reasonably up-to-date Mac Pro for more years than not in the past decade.
I think a lower starting price is possible even with the Mac Studio, but I wouldn't expect it to be any lower than $5K, and like the 7,1 you'd have to put $1K into it to get a reasonable configuration for most people (bumping the storage, GPU, and RAM costs you $800 right there, and back when it shipped with a 256GB SSD and the comparatively much worse 580X you'd probably have wanted even more, even if you did things like RAM aftermarket.)
Cutting your expenses to be less than your revenue equals profit, and that's how Apple had its first profitable quarters under Jobs, despite their market share continuing to erode. They only substantially improved on their actual Mac sales and market share year-over-year with the iMac, and when they talked about their results they didn't mention their professional customers, they mentioned the iMac. Had they not released that computer, cost savings or growing their other computers wouldn't have mattered, they couldn't have become the company that dominated the 2000s.
Think Different was entirely about rebranding Apple inside and out, not trying to directly sell Macs (hence why it featured no products). From Jobs himself: "We have to prove that Apple is still alive [...] and that it still stands for something special" (to TBWA\Chiat\Day ad man Lee Clow.) "It was directed not only at potential customers, but also at Apple's own employees." (Isaacson, 328). It was an expression of Jobs' personal philosophy, not trying to speak directly to professional Mac users.
The idea of the Mac Pro as a Halo car has been brought up a lot, but if that were true I'd argue Apple would clearly be in much worse shape now than it is, given that there hasn't been a reasonably up-to-date Mac Pro for more years than not in the past decade.
You can cut all expenses to zero, if you have no income, you have nothing. You need income. Yes, cutting expenses helped, but without the income part, you have NOTHING. You must SELL.
The think different campaign was great, and proving you have survived was a life line appeal. PLEASE STILL BUY OUR STUFF. The appeal was made to the crazy ones. The ones that stuck around. And the halo crazy ones had an outsized effect back when apple needed to lean on these die hard lunatics.
Maybe Apple will never need to lean on those users again, but burning that bridge is stupid IMO. Most other companies would kill to have such insane (and well heeled) loyal customers. Instead, Apple is basically acting like they dont mind pissing in our cheerios these days. No bueno.
Furthermore, Porsche's bread and butter car is their bloated sack of crap Cayenne SUV for moms/dads that want to lie to themselves about how 'sporty' they are and flash the Porsche badge in one-up'ing the joneses. They would make no sales of it without still making 911s (and the occasional 918). The Cayenne is made special because Porsche is seen as a performance leader because of its halo products.
If you disagree, fair enough, we disagree.
There are reasons for that.
"Although your guests are supplying the wine, you're supplying the service.
Former beverage director Sean Park says, “Diners may whine and groan about the upcharge, but the reality of proper wine service is that there’s so much more to it than simply pouring a glass from a bottle.”
Your servers will open the bottle, bring it to temperature, serve in the appropriate stemware, and refill as necessary. For an exceptional bottle, you want these rituals done right so the service is worth paying for.
Recouping on beverage costs is reason number two. It may sound trivial, but in the low-margin restaurant game, every dollar counts.
On average, turn times for tables that order wine are longer than turn times for tables that don’t. For that reason, corkage fees encourage diners to only bring a bottle if it’s truly special. No diner wants to spend $35 to serve a $15 bottle, after all.
The third reason for a corkage fee is that many restaurants are trying to help mitigate investments they’ve made in their wine program. Curating an impressive wine list, keeping a sommelier on staff, and training staff on your wine selection are pillars of a reputable program, critical to attracting wine enthusiasts, and expensive.
A.J. Bruno, a sommelier with 10 years experience, explains, “[Guests] often don’t understand that sommeliers painstakingly search for wine that will pair with the food presented by the chef and that it is priced in a way to keep the lights on, rent paid, and hopefully make a small profit as well.”
Charging a corkage fee allows restaurants to give wine enthusiasts the option to bring their own bottle without undercutting the expenses they’re incurring."
In the MacPro case pricing may be high to keep Apples' higher profit margins on a smaller volume product which has high relative development and production costs.
The 128GB 20-core Ultra Studio starts at AU$8799. The 32GB, 8 core, W5500 Intel Mac Pro starts at AUD$9999.
I don't see that stratification changing.
I would also not bet on any eGPU access for current AS Macs, if the AS Mac Pro keeps PCI / MPX GPUs, including compute-only versions. That will arrive, if it does, with revisions released AFTER the Mac Pro.
Because Apple Silicon has not produced price drops on machines where it's replaced Intel.
Because 2019 didn't produce a cheaper Mac Pro when dropping one of the 2013's GPUs, or dropping the iMac Pro's display.
Because the price of the 2019 is not the processor, it's the case.
Because Apple has a captive audience, and they can.
Because Tim needs to make a line go up.
Because slots are a premium feature, a significant, and expensive premium feature.
Because a Mac Pro that is slower than a Studio just makes the Mac Pro look slow, rather than making the Mac Studio look fast.
Because the Apple Silicon Pro & Max are easily outperformed by consumer Intel and AMD systems.
Yes, it's more than enough for users, but the Studio is designed to cover every user class the old Intel Mini and 27" iMac covered. The Max is a laptop chip, not a workstation chip.
To paraphrase: if you need professional expansion, you need the highest-end professional cpu, so your system is "balanced".
You are more or less explaining back to me what I wrote in my own post. I pointed out the way things are and said I feel it should work differently. I didn't express bewilderment as to why it works the way it does.
You spend some time explaining the "corking fee". I think that is pretty American, but perhaps not exclusively. I don't think I've ever come across that in Europe. Might still exist, though.
I'm talking about bottles that you buy from the restaurants' "cellars", or kitchens rather. They buy a complete product that they just re-sell. But instead of adding typical "middle-man" upcharges they often add close to 100% of the purchase price.
As you mention, there is a wide range of wine knowledge and quality in restaurants. It's important to recognize that a guy pouring wine isn't a sommelier. Having a few sommeliers among my friends, I know what it takes to get certified.
This brought a smile to my face. I'm sure there are authority-seeking individuals who are nodding their heads in agreement with this. I'm not one of them. I think that's the most polite way I can respond to that.
Someone wrote something similar further up. My suggestion was not about making less money. It was about charging for what the restaurant is actually responsible for: location, ambiance, menu, craft (wannabe sommeliers don't count), and so on.
This sounds like a reiteration of my point: Apple should charge for their R&D, designs, and craft—but not so much for just reselling stuff from others. In a zero-sum game, this would lead to higher base prices and lower prices for off-the-shelves 3rd party components. But in the end, I want base prices to stay the way they are, and for Apple to just lower RAM and storage prices.
There are many phenomena in the world that are interesting to discuss, but it's off-topic so I don't want to dig into it.
I have a theory: could RAM be added via MPX RAM add-on modules? CXL style.
Another theory: what if the dGPUs themselves are also a way to add RAM (shared memory)?
Main RAM of the M2 Ultra: 192GB
4 x “Lifuka” dGPU MPX modules: 4 x 256GB
Total shared memory: 1.2TB (of which 192GB at the lowest latency and a far pool of 1024GB at higher latency)
Costly to pay for unwanted dGPUs just to get more RAM but at least there would be a way for those high end users to do it.
Am I talking nonsense?
Another theory: what if the dGPUs themselves are also a way to add RAM (shared memory)?
Main RAM of the M2 Ultra: 192GB
4 x “Lifuka” dGPU MPX modules: 4 x 256GB
Total shared memory: 1.2TB (of which 192GB at the lowest latency and a far pool of 1024GB at higher latency)
Costly to pay for unwanted dGPUs just to get more RAM but at least there would be a way for those high end users to do it.
Am I talking nonsense?
I think Apple wants to make Mac Pros as attractive as possible for people to buy. I think using Apple Silicon across the board is a way to achieve that.
So far we've seen the replacement of Intel's cheaper CPUs. Let's see what happens with Xeons.
I am not of the opinion that Apple tries to "rip off their customers for as much as they can".
The Mac Pro isn't a "fast computer". I also don't see it as a luxury computer. The Mac Pro isn't an expensive computer, per definition or by nature, even if it ends up that way.
I'm not into PCs anymore and don't look at benchmarks that aren't relevant to me, but this has not been my feeling (or are you taking GPUs into account here?).
I fail to see the logic. If you need expansion, you need expansion, right, not a high-end CPU? And if you specifically go for a high-end CPU, that would typically lead to a more imbalanced system, unless you bring everything else up too, which wasn't specified.
Anyway... I feel I'm increasingly spending time on tangents. I think I'll rest my case until March and only reply to direct quotes.
I can understand the whole "keep the old chassis to transition to new internals" idea, but I really hope Apple will give the (hopefully) 2025 M3 Extreme Mac Pro chassis a redesign... ;^p
what characterizes xeons (and ecc memory) is that they keep their frequencies, even if they work non-stop for several days. I have read testimonials about Threadrippers not having this stability. It's the same with pro graphics cards (with ecc memory).
What makes the mac pro "pro", it is unfailing in calculations. It cannot be compared to core i and ryzen, which are very suitable for a large number of works done by certain professionals.
If Apple wants to keep a mac pro, it's to make it a reliable machine for editing its Apple tv series, for rendering, special effects and 3D pixar effects, for composing movie music with virtual orchestra , without an unexpected crash preventing operations.
What makes the mac pro "pro", it is unfailing in calculations. It cannot be compared to core i and ryzen, which are very suitable for a large number of works done by certain professionals.
If Apple wants to keep a mac pro, it's to make it a reliable machine for editing its Apple tv series, for rendering, special effects and 3D pixar effects, for composing movie music with virtual orchestra , without an unexpected crash preventing operations.
Generally if you get a bit-flip while you're doing something like 3D rendering, it doesn't really matter, because it will just appear as noise and eventually all noise converges towards the truth. Same with training an AI model, eventually that blip will just get drowned out by correct data.
If you get a bit-flip while doing a scientific or mathematical calculation, your entire calculation could be garbage that needs to be thrown out.
There are times when ECC RAM is absolutely necessary, and other times when it isn't. There's definitely no harm in having ECC RAM though, apart from cost.
The biggest problem with the Mac Pro is defining what a pro machine is. It's like trying to define a "pro car" - is it a supercar, a truck, a Maybach, or a tractor? With PCs you can build exactly what you want, but with the Mac Pro Apple have to try to make one machine that is going to please everyone, and that's an incredibly difficult, probably impossible task.
If you get a bit-flip while doing a scientific or mathematical calculation, your entire calculation could be garbage that needs to be thrown out.
There are times when ECC RAM is absolutely necessary, and other times when it isn't. There's definitely no harm in having ECC RAM though, apart from cost.
The biggest problem with the Mac Pro is defining what a pro machine is. It's like trying to define a "pro car" - is it a supercar, a truck, a Maybach, or a tractor? With PCs you can build exactly what you want, but with the Mac Pro Apple have to try to make one machine that is going to please everyone, and that's an incredibly difficult, probably impossible task.
You will need to get the M2 wheels, the intel ones will not be compatible.
Last edited:
According to me :
- mac pro with xeons (or a windows/linux workstation) : a long freight train (slow but carries a lot of freight without weakness)
- pc with core i9 or ryzen 9 : a powerful sedan
- mac pro with xeons (or a windows/linux workstation) : a long freight train (slow but carries a lot of freight without weakness)
- pc with core i9 or ryzen 9 : a powerful sedan