Sure, and I appreciate all the work that has gone into it, but I'd prefer that Apple release a Mojave patch to address the vulnerability.
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 1,1-5,1 What machine do you expect to be using in 2025?
- Thread starter mode11
- Start date
- Sort by reaction score
Unfortunately, I count about 17 different security updates (this includes Safari updates; see the Apple security updates page) that have been released for Catalina since Mojave went out of support in September 2021. The likelihood that Apple will fix just this one newer problem for Mojave is approximately zero.
I do wish Apple had longer support life cycles for their software, but wishing for a thing does not make it so.
Hence one of the reasons for this thread.
As I said, I will have my eye on a high-end Mac Studio a few years from now, but I'd be glad to consider any Mac Pro system if it makes sense for me (and if my wallet would forgive me). Until then, I'm reasonably happy with my maxed-out 2018 Intel Mac mini, though it does fall short sometimes.
I also expect to keep the Intel mini even after moving to an Apple Silicon Mac. I occasionally need to run very old versions of Windows in VMs and I doubt this will ever be possible on an AS system. (I even have a PowerPC G4 Mac mini that I keep around for old Mac software, but I don't use this system very often.)
How does the upgradability of the 7,1 compare to that of the cMP? I'm asking not so much about built in number of slots / ports but rather any limitations inherent in the 7,1? For example upgrading the internal SSD wasn't possible when it was first released. It wasn't until Apple decided to make it possible. To my knowledge no such issue exists with the cMP series. Aside from the internal SSD restriction does the 7,1 have any other similar restrictions? Also I am not referring to OS drivers (though those are important) but rather hardware restrictions built into the 7,1.
^^^^What you are listing as a limitation is not a limitation. The cMP did not have a T2 protected internal drive therefore the stock OEM drive could be replaced or added too by any number of 3rd party or Apple drives in different configurations. The T2 OEM drive in the NcMP can only be replaced with Apple Branded SSDs and only in a particular configuration and only with another Mac present to assist.
Another restriction that applies to the cMP. Apple and Nvidia had some sort of disagreement and Apple stopped including drivers for Nvidia GPUs after High Sierra.
The cMP will also allow 3.5" HDs to be installed internally. The NcMP will not.
The cMP allows for an internal Optical drive, the NcMP does not.
As far as ports, PCI slots, and expansion capabilities the NcMP has that on spades over the cMP. Just read the specs to see the differences.
Lou
Another restriction that applies to the cMP. Apple and Nvidia had some sort of disagreement and Apple stopped including drivers for Nvidia GPUs after High Sierra.
The cMP will also allow 3.5" HDs to be installed internally. The NcMP will not.
The cMP allows for an internal Optical drive, the NcMP does not.
As far as ports, PCI slots, and expansion capabilities the NcMP has that on spades over the cMP. Just read the specs to see the differences.
Lou
Maybe I am missing something but how can it not be a limitation when the 7,1 can only work with Apple branded SSDs and only in particular configurations. That sounds like a limitation to me.
I voted for the Mac Pro because it's what I want. Ultimately it will come down to the total cost of the 8,1 MP vs a Mac Studio w/necessary expansion peripherals. October is just around the corner, finally....
That's not what I said. The T2 can only be replaced with an Apple branded (actually a pair) SSD. But, the NcMP can "work" with any SSD either mounted on a PCI card or in an internal cage. My NcMP has ten internal SSDs. The OEM T2 unit, six blade SSDs on two PCI cards and three 2.5" units mounted in a cage. My boot drive is on one of the blades. The T2 is only there in case of catastrophic failure.
Lou
Latest upgradable Mac Pro, I will not go into Studio or similar locked direction.
I believe Apple will continue to produce a solution for music and video industry. The market is too big to neglect it. I remember seeing on Youtube people comparing setups between two 2013 cMPs with an army of external cards attached to, versus a clean 2019 cMP. I'm waiting for 2023 to see if a refreshed Intel 8,1 is announced along with the Mx variant. I will decide then what direction to take, based on upgradability.
Last edited:
Neil Parfitt, who's been probably the poster child for why the 2019 needs to be what it is - a generic slotbox, on the failure of trying to use a top-of-the-range Mac Studio in the role.
LOL. At 3:08 in the video he says he expected it to work. At around 5:08 he says that if he didn't have old legacy tracks from previous versions of LogicX and was starting over from scratch ... again he might try the Studio path .
Slot has diddly poo-poo to do with what he is talking about. Nothing. What he is talking about is that if you have data tied to older applications and older macOS versions then it is more painless to buy more of the same 'old stuff' you already have.
Another comment he makes in the video is something to the effect was ' the Studio is a Mac for people who want to go forward'. That really is not really restricted to the Studio in the general case. All Macs are for folks who want to go forward. Apple is not a super fan of oldest as possible. 32-bit apps? Gone. Kernel Extensions ? on 'death row' (deprecated). Security upgrades for macOS version n - 4 ... poof, discontinued. GPU driver for you 3rd party GPU card in macOS on M-series? dead in the water. (slots without drivers for the hardware in those slots is not a functional solution. Gesturing wildly at the slot being some magical 'get out jail free' card is gross misdirection from the real core issue. )
That issue existed with the intel macs also. If Apple came out with a new W-3300 Mac Pro in November with all the slots of the 2019 one , it would not run macOS 10.15 (Catalina) [ nor 11 (Big Sur) nor 12 (Monterey) ). So a 2023 Apple Silicon Mac Pro that can't run 11 or 12 is just the standard modus operandi for Apple; not some new M-series policy. A new M-series Mac Pro with a new SoC that required updates from the audio plug-in software would have all the same hiccups the Studio had ( where the audio software to trying to cling as far back as possible to hardware and are not putting most effort into Apple's leading edge hardware and OS (and/or waiting for Apple to fix bugs in new OS. . ). Slot or not would make no material difference at all. In fact, if required new DriverKit drivers that had 'growing pain' quirks , it could actually be more unstable than the Studio.
Very similar issue with LogicX. It has a sliding windows of macOS versions that the newest version will support.
So if hitching your sound project to a plug-in that is dead (no updates , no support ) and permanently 'stuck in time' then have real issues regardless of architecture change.
What Parfitt is very illustrative of is the baseline reason why Apple is in no huge hurry to update the Mac Pro. And a decent chance continue to sell the aging 2019 model for perhaps another year or so even after have a new one. There is decent sized market for selling very expensive old stuff that isn't bleeding edge performance.
Those are not SSDs. There is only one Apple SSD in the Mac Pro 2019. What Apple somewhat confusingly sells is SSD modules. 'SSD' there is an adjective , not a noun. It is what type of module, not what it is. Those cards are basically NAND daughtercard modules. There is no SSD there are at all. You can't have a SSD without a SSD controller (the 'brains' of the SSD).
What Apple has done is decompose what is typically just one single drive implementation board into three parts. SSD controller in the T2. NAND daughter cards holding the actual storage modules. This gives them some flexibility to repair a SSD where the NAND have catastrophically failed without having to replace the whole , very expensive, logical board. The so called 'upgrade' path is purely just a side effect of that mechanism for repair of the drive.
If pop in two new modules you'll need a 2nd pack to run Apple Configurator to restore the SSD's function because you have basically 'blow up' the old SSD drive.
Those are repair parts. Apple vintage/obsolete policy puts a window on those parts availability.
Because their NAND modules have to be initialized when you put them in, it is also a very dubious idea to put old ones into a T2 mac that might accept them. If the initialization process wipes out all the metadata that the SSD controller was using to map out bad NAND cells and do wear leveling accounting, what you will have is a drive that is more likely to fail badly in the future ( the older the replacement module the more likely will have a larger wear leveling policy disconnect). So the notion "when Apple stops selling new replacement parts, we'll just go to the old machine boneyard and toss in old used parts" is a dubious path. How many folks go out and buy a 10 year old used SSD to put into their Mac as a replacement part now? Not many.
Past Apple's window of support for the T2 SSD drive , the viable work around is to just do all the booting/work on another internal SSD. If don't use the T2 SSD drive much it is very unlikely going to fail. But any major system reset is going to require that T2 SSD to be working. Going to at least need to have a "maintenance macOS" instance there that you periodically use to do some work with. So that is a substantive difference.
An out of the box, BTO configuration? No. With a J2i bracket installed Yes. Apple isn't a fan of SATA drives. There happened to be an empty void downstream from the CPU heatsink, so Apple tossed a bone there. It changes to the case happen to remove that void (or some other large enough empty space) then I wouldn't count on Apple keeping it in a future Mac Pro. At "half the volume" of the current Mac Pro there is likely still extra room to toss at least one drive in. And making a Mac Pro about the same size as the Studio doesn't make much sense. They need some product differentiation. Even more so since Mac Pro likely will have a "Ultra" class SoC also.
Apple had a large hand in creating the "internal time machine drive" paradigm as a very common set up. Apple preaches the safety of multiple backups and then take away a near-line internal backup. People are just going to complain about that.
The 7,1 isn't going to allow you to hack the core UEFI firmware. Part of the T2's job is to validate that the core firmware has not been hacked. The boot up process starts on the T2 and then the Intel CPU is handed a copy of the validated firmware.
There are some shenigans can do on top of the UEFI layer after the Intel CPU is running. However, not going to do things where play the "screw Apple I'm doing x,y.z with the firmware" kinds of 'upgrades'.
Very Long term the 7,1 is going to be stuck with PCI-e v3. When PCI-e v6, CLX 3.0 hardware comes along ... it is not going to work. There is going to be a different protocol on top of PHYS PCI-e that is going to come into play over next 5-6 years that did not happen over last 10-15 years.
Similarly for the evolution for PCI-e v5-6, some cards will be shifting back to lower lane counts. (e.g, x16 -> x8 or x8 to x4 ). That isn't going to work well with the 7,1 either. The slower lane speed will be halved again by halving the lane count. ( similar issue with getting PCI-e v2 x16 cards to work well in a x4 PCI-e v3 slot. For example, the AMD 6500XT only has x4 PCI-e v4 ). Once get to the point in more distant future where PCI-e card vendors are only targeting PCI-e v4 as the minimal baseline the lane shrinkage will be more acute. A substantive move to fewer lanes largely did not happen over the last 10-15 years.
The new Intel GPU cards don't work well without Resizable Bar. (ReBAR). Similar to more CXL rolling out in workstation ecosystem over time is that GPUs that do not create a hybrid "Unified memory" are likely going to have problems over the long term. Short term, Intel got the timing and drivers wrong, but, in the long term, the notion isn't wrong. ( Graphic APIs allowing GPU to do direct bulk data pulls from extremely fast SSD is only going to drive up bulk data traffic to GPUs over time. )
Current (and older) gamings are constructed around the premise that the data path between CPU and GPU is relatively slow and bandwidth/latency bottlenecked. Apps optimized 5 years from not probably won't be laboring under that same assumption. So the next 10 years software is likely not going to look completely like that last 20 years of software.
The 7,1 being a larger container for older and backwards looking cards is somewhat different than the core notion of "upgrading" (moving forward) . The 7,1 won't be different in ability to collect older , backward looking stuff.
But the demands/requirements of future hardware is likely to change over the next 5 years. Next 2 not so much, but by 5 there will likely be some substantive shifts.
Just saw new entry on Macrumors front page.
Parfitt also owns a rack mounted Mac Pro, not the tower version.
"Nobody" wants a racked Studio ....... so Sonnet goes and creates this:
Sonnet's Latest Enclosures Help You Rack Mount Your Mac Studio
Professional storage and networking accessory company Sonnet Technologies today announced a pair of solutions for Mac Studio users who are looking to...
www.macrumors.com
because "nobody" is going to buy it. A couple of old audio cards (presuming have working drivers ) new system. Less vertical rack units ( 3U ) consumed than a Mac Pro ( 5U ). Up to 3 HDX Avid Pro cards in 3U won't work for everybody, but it will work for way more than just a few.
A Studio Ultra 20C/48G , 128GB RAM , 2TB SSD $5,200
Three slot Echo III rack $1,700
$6,900
Rack Mac Pro 16C , 96GB RAM , 512GB SSD , 6600X GPU $9,800
More slots but more rack U units also. $3K more expensive (plenty of budget that could be used for the "sneaker net" Thunderbolt external SSD system that Parfitt is talking about) . And on software DAW workloads with working software ... slower.
A rack mounted Mac Pro with Ultra 20C/48G , 128GB RAM , 2TB SSD , and 5-6 slots ( perhaps +3 add on with a cleaner solution than here ( internal PCI-e socket to expansion board to fill out the standard rack width ) for $9,800 would be way more competitive over long term. Tower version might shrink on slots but the rack need to "bulk out" to fit the standard rack width anyway.
Now 3rd party GPU drivers here at all and yet a very viable product for Sonnet Tech to create.
Edit: P.S. on Sonnet site they have a section called "Road Trip Ready".
xMac Studio - SONNETTECH
3U enclosure to install and secure one Mac Studio with optional PCIe card expansion in a standard 19-inch rack.
That has some overlap with the use case that Parfitt was actually talking about. (how to move from place to place and still have access to your DAW). There are downside to dragging around 40 lbs of chassis when moving around. ( part of why Apple is charging $100/ wheel.
)
Last edited:
That's the guy who was demonstrating how much better it is his 2019 Mac Pro (rack mount style), versus the 2013 Mac Pro!
That. To me it makes more sense to just buy a refurbished 7,1, if they don't make happen the Intel 8,1. I'm never going to look for staggering performance, I prefer long supported software upgrades, which can be extended with OpenCore. Like @tsialex mentioned in another thread, you can buy later a pair of CPUs for a low price.
The Ultra SoC can decode and encode more 8K video streams in near real time than a MP 2019 ... how are they avoiding the video industry? That is basically provisioned by making a trade-off on GPU+RAM modularity.
As for the audio industry. See posts above. It isn't uncovered even before they release a Mac Pro
They should have a Mac Pro with some ( > 2 ) slots. But it also likely not going to be the 40lbs monster tower either. The more Apple narrows down the Mac Pro to a smaller set of specific industries then the more likely just going to get SoC bundled specialized accelerators that a just better than many of the other options.
3-4 slots would give them something that was useful in a broads set of use cases than just music and video.
( 5,1 did OK with one slot dedicated to GPU, so 3 can certainly work. )
Again the Ultra SoC , let alone going to a Quad solution (40 cores) , takes care of covering the build of that "two 2013 MP" need. ( which was more so short on CPU compute limit of those 12 cores , than either GPU or to a lessor extent RAM. ) Modular GPUs had nothing to do with that. ( the slots that GPU(s) took up may not may the transition. )
Very unlikely that post end of 2022 going to get a new Intel model. The transition was suppose to be done inside of 2022. Even if Apple had a "Intel 8,1" penciled in in their 2018-2019 era roadmap , it is probably scrapped at this point. Pandemic delays , Intel fumbling the W-3300 as a high volume product , etc. Once Apple covered the iMac 27" part of the line up, all of the major volume sellers were converted. If they cover the last Intel Mini this Fall even bigger "door close". There is no volume to support macOS on Intel development over a very long term.
There was a window back in 2021 where the W-3300 would have been reasonably competitive for a while. At the end of 2022 , it is not. Dell, HP, Lenovo ... all not touching it. So why should Apple? W-3400 is an even bigger mismatch to macOS thread limit than the W-3300 is. So again not buying Apple much. Intel "mainstream workstation" (Fishawk Falls) ... not even done with validation. Also not going to arrive until well after end of 2022.
As "side by side" sales path for Apple at this point is very little just selling the 2019 model into the future. Its basic R&D is paid for. A cheap upgrade could be tossing some new AMD GPUs at it, but that would be about it. If all that Apple is doing is selling the 'hyper modularity" aspect rather than performance .. it has it. Either you pay for the hyper modularity or you don't. Apple is making their money elsewhere. Apple stretched the MP 2013 for 6 years. They could stretch the 2019 model for 4-4.5 years easy.
I agree, that's what I'm interested into. But I have a feeling the new Mac Pro will be a new 6,1.
There is large mythos around paying more for a Mac Pro gets you longer software upgrades because of the hardware. That is largely not true. Mac Pro has eeks out longer coverage along two lines. One , for over the last decade, Apple has updated Mac Pro generally slower than most other Macs. That has the side effect of extending support only because the Mac Pro don't get superseded at a "normal" (for Apple) pace. The hardware isn't doing that. Apple sloth is doing that. Paying Apple more doesn't 'buy' more (or less) coverage. It is just not a factor in how they approach vintage/obsolete status.
The second path has been ... we'll just hack around Apple's lack of support. That you have to do a hack technically means that is not supported. If Apple was doing the work , then you would not need the hack. That isn't support. That is more about control ( "screw Apple I'm going to do what I want to do"). 5 years from now I wouldn't make major bets on that.
First, as I mentioned in previous post the firmware is going to be locked down even tigher than the T2 systems.
Second, macOS is now off is a completely separate whole partition where only Apple signatures work. The security layer there is likely only to get even tighter and deeper 5 years from now. (even more so after all the pre T2 system roll off the supported list).
Third, kernel extension are going to be removed at some point. Apple deprecated them in 2019 ( a year before the 2020 transition). They are already three years on the deprecated list. Thinking that is going to last another 3-4 years is loopy. So all the hacks that depend upon kernel extension may disappear in next 3 or so years. Pretty likely a more secure kernel shows up when kernel extensions are permanently dropped. Decent chance that more secure kernel is going to have a much more secure handoff between the T2 (or better) security processor and macOS.
At which point, OpenCore might run older and non-Apple stuff but would have far more problems with newer macOS versions. There is probably a camp that is going to rely on the T2 being old in several years and they whack away at the unsupported security processor until they crack and subvert it. Perhaps. Or Perhaps not.
pair of CPUs? Sure for a 5,1. For a 7,1, I'm not sure what the high utility for that would be for that. In the future buy a cheaper older single one. Sure. However, that isn't going to extend Apple's support horizon for the system.
In fact the more folks who buy used/second mac Intel system only further weakens the funding support poll for Apple upgrades for macOS on Intel. ( only going to collapse faster if that is a larger segment. ).
There was a decent rumor that there was a "M1" Mac Pro with one slot. M1 makes it somewhat of a dead end because Apple announced the M1 line was over ( doesn't mean they didn't make a prototype that will never ship in relatively large numbers ).
Technically a one slot isn't a slide to a 6,1. If they left the J2i bracket ability and/or multiple M.2 / U.2 / E3 EDSFF slots for more internal storage that would remove a large chunk of the 6,1's issues. Need more than one and only one internal drive. If not using one of the standard PCI-e slots for increase internal drives than just one more substantively better.
But it wouldn't be a huge stretch for some configuration of that one slot wonder logic board to turn into a multiple slot , rack unit that just routed the slot expansion through that one slot internally.
I really don't think it is going to be a repeat of the 6,1 because Apple has already largely shipped something better than a 6,1: Mac Studio. Similar issues. One and only one internal hard drive. Overly leans on Thunderbolt to solve its painted into a corner I/O problems. The 'double" GPU of the Ultra SoC some software is taking advantage of and other , unoptimized , stuff is not. Not completely the same. Better manage not to paint themselves into a thermal corner.
So a Mac Pro being "Just bigger Studio" chassis... I don't see how they will get any decent product differentiation. More TB sockets isn't going to buy much, if anything. If don't uncorked the single internal drive problem then sitting in the same 'ditch'. APFS pragmatically is useful on SSDs. The cheapest path to add a wide variety of SSDs internally would be a generic PCI-e v4 (or better) slot connector. ( a specialized storage connector M.2 would be more volume effective , but "cheaper for Apple" ? Just push the specific storage connector cost to the end user. )
Toss in the internal SATA/USB sockets the current 7,1 (2019) has and a J2i bracket and even has spinning platter storage without Apple having to bother (or research ) much. Apple could select and put a discrete SATA controller on the logic board and slap a high mark up on it ( to push a price gap away from the Studio). As much as being a container also helps them drive the system price much higher I suspect they are going to buy into that. ( it is just fatter profits.)
It is certainly not going to be like the 6,1 (2013) price point. Those started at $2,999 (which again the Studio has already covered). The Ultra powered Studio's start at $3,999 . The tower Mac Pro starts at $5,999. They can bump the minimal configuration RAM and SSD capacity. But that is a big gap to fill with only playing with BTO options. They need some substantively different value-add features to lift the price point. Slots and storage should work without having to lean 'too hard' on overpriced BTO options to get there.
If Apple was super lazy there is likely too many TB controllers in their SoC solution. The extra cheapo path would be to just take 1-4 discrete TB peripheral controllers and solder those to the logic board with a single physical slot socket attached. I'm more worried about something spectacularly lame like that than to a repeat of the 6,1. That they will only use laptop optimized dies to effect the solution. That is what is likely going to drive a one-slot-wonder. They need an augment die that gets them some real desktop workstation worthy I/O coming out of the SoC. It will just be Sonnet Tech's solution in a more expensive box with Apple Industrial design sprinkled on top. Mainly just get Apple 'Sherlocking' Sonnet Tech's effort.
Yup, and if you check that video you'll notice that apart from basic compatibility, the top of the range Ultra was dropping notes, because it's incapable of performing the job as well as the 2019 machine.
For whatever reason, it's not up to the task.
Anyone thinking a professional machine can get by without being able to maintain access to historical working files, is so deep in dunning-kruger land you can ignore them.
When a large government lab or organisation upgrades and throws out its old IT tech, you will find guys crawling through the dumpster for spares. When the department can no longer read their old files, those people that saved the old tech will be hired to read & convert them !
That was my understanding also and that’s why I’m inclined to purchase an Intel model. I believe we will be able to take an educated decision soon, once we know what choices we have for Mac Pro.
I’m using my 5,1 for 10 years, thanks to OpenCore. I believe we can achieve similar results with a 7,1.
So, you are really expecting that Apple will still send macOS releases for Intel processors at least to 2027? I really think that you have to lower your expectations, by a lot.
People betted against me that MacPro6,1 would be fully supported up to 2025 and I always warned that didn't made any sense. I won't expect any new macOS releases for Intel processor later than 2025, five years since the last Intel Mac (2020 27" iMac) was released and I'm being optimist here. Since Apple Silicon, Apple is currently dropping support much faster than expected.