But did Apple actually commission GPU manufacture under licence from AMD, using a smaller process node? First time I've heard of this, though I wasn't following closely. Surely the GPUs were off-the-shelf from AMD? The only custom stuff I recall is the better iGPUs Intel used for Mac laptop CPUs.
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
I don't know about Apple having AMD GPUs made on a smaller node just for them, but the PCB for the MPX GPUs was definitely custom to Apple, especially the dual chip johnnies...?
IIRC there were some Vegas in MacBook Pros that did not match what was available on PCs, and were manufactured on a different process. No idea if Apple produced them, but they weren't off the shelf GPUs.
So if TB5 ports can clear up bandwidth issues with non-GPU PCIe cards (in a TB5 PCIe expansion chassis, of course)maybe we see Apple implement the following headless desktop/workstation line-up:
Mac Pro Cube
- Mn Mac mini
- Mn Pro Mac mini
- Mn Max Mac Studio
- Mn Ultra Mac Studio
- Mn Extreme (Quadra...?) Mac Pro Cube
- Mac Pro TB5 PCIe expansion chassis (mini-tower form factor w/optional rackmount kit)
Mac Pro Cube
- 64-core CPU (48P/16E)
- 960-core GPU
- 192-core NPU
- 2TB LPDDR5X ECC RAM
- 64TB SSD
It's interesting, my feeling is I'm fine with both MacOS and iOS to a point where I don't even think about them really. They're both extremely easy to use and efficient. My only issues are I wish there was a better folder structure built into the phone. Other than that I can't really think of anything that bothers me lol. But I'm SUPER excited about WWDC and the glasses could end up being awesome.
Yep. I like the Real glasses. Nice big movie theater sized screen and very comfortable. Not quite BIG SCREEN level of immersion due to the lack of full eye enclosure and locations "virtual theaters and whatnot", but still very cool.
Yeah, I think MacOS's main issue is that Apple seems to feel it can't be "boring", and it's getting hitched to the iOS release schedule. Going back to major point updates every 2–3 years would, I think, make most end users happy (I know of no one who likes updating software, and certainly for professional use cases one ends up clutching to rock-solid HW/SW configurations and one is usually loath to update.)
well pcie 4.0 at X4 max is not really that much. Why not have ext PCI-e ports as an expansion chassis does not need video out getting in the way and on the higher end cpu's there may be unused pci-e lanes or maybe some TB out's that chips video gpu does not have the power to feed.
I think the issue is that macOS and iOS are so unified and intertwined now in the Apple Silicon era that they sort of need to be hitched together.
Most SDKs are shared nowadays, and it would suck if macOS was stuck on an older version of SwiftUI or if the GPU was hindered by only having Metal 2 instead of 3.
Not to mention you can run iOS apps on macOS, and people expect that apps look and run identically on macOS 13 as they do on iOS 16.
Finally you have drivers and cross-platform stuff like sidecar and hand-off that really need to be running the same version.
Sure, of course MPX cards are custom built for Apple. I was questioning the specific claims of AMD providing Apple with custom GPUs, and of licensing Apple to fab their designs on a smaller process node. Both of which sound highly dubious (particularly the latter).
I actually just had a really radical thought reading through this thread just now....
What if with the glasses, apple's main long term goal is to ditch support for monitors/screens altogether and force users to buy/use the glasses as the only supported video output device/monitor/screen?
Think very long term with this radicalization though, not implemented tomorrow....
Think about it -- this way, they develop both macOS and iOS to output the same, thereby reducing macOS code base (simplifying and making it even more "efficient") doing the same to iOS, and as mentioned so that all apps, whether ran on macOS, or iOS, output the same exact thing/provide the same user experience -- "regardless of device you are using".
Pretty crazy idea huh? -- Thoughts?
What if with the glasses, apple's main long term goal is to ditch support for monitors/screens altogether and force users to buy/use the glasses as the only supported video output device/monitor/screen?
Think very long term with this radicalization though, not implemented tomorrow....
Think about it -- this way, they develop both macOS and iOS to output the same, thereby reducing macOS code base (simplifying and making it even more "efficient") doing the same to iOS, and as mentioned so that all apps, whether ran on macOS, or iOS, output the same exact thing/provide the same user experience -- "regardless of device you are using".
Pretty crazy idea huh? -- Thoughts?
Productivity is a huge possibility. Think about drawing a screen on on your wall and it stays pinned there. You can have “stationary” screens anywhere you want augmented into your physical spaces. Hand gestures for virtual keyboards, mice, and actual UI hands.
Some of this exists already but the UI has been poop, but promising. Apple could make it good and actualized.
While I doubt they would make it the only screen, at some point your glasses may be the preferred screen. For everything.
Think of the wifi of video And UI Input protocol. So anywhere you go a computer system will give you a connection and your preferred UI will pop up to interact with that system. Go to a bank. A hotel. A grocery.
I don't think Apple has any long term goal with the glasses, honestly. And I don't think that's particularly a bad thing. Assuming most of the leaks are in the ballpark, they seem to be positioning the glasses as an expensive "it can do anything!" type of device, and will likely do the usual recent Apple thing of iterating and refining based on what consumers do with it. Remember when one plank of the Apple Watch pitch was watch-to-watch social engagement? That got sidelined real quickly to focus on the health and fitness and notification features that people actually used versus digital touch.
Being able to stick infinite screens wherever you want has some potential future utility, but like a lot of AR stuff the pie-in-the-sky future stuff would require A) everyone wear glasses that B) share a significant amount of interoperability to function at that level. I don't think displays are going anywhere anytime soon, especially since there are much more stringent limits on what you can do with virtual screens right now in terms of horsepower and hardware.
In a few decades? Who knows? But I don't think as a publicly traded company Apple could ever afford to try and plot a line like that.
Really?
The most valuable company in the world, with a massive stockpile of cash, and major market influence can't afford to try that? 🤣
If anything, apple is the ONLY company that can literally AFFORD to try something like that...
Last edited:
The glasses are going to be a facetime version of the Jedi Council meetings from those awful Star Wars prequels.
transparent people sitting on empty couches, that you can facetime with, and even worse, they're going to be infantalised memoji versions of people - because sending animation data of a model is lower bandwidth than sending video.
There's no technology path on the horizon for opaque-capable eyescreens that aren't based on passthrough video, and retina-quality passthrough video isn't going to be in a non-specialist-bulk device any time soon.
Welders can wear a welding helmet all day to do welding, office workers won't to do office work. Likewise, 3D modellers / specialist media producers can wear a bulky HMD. General computing won't.
I personally thought it was more likely that Apple's "glasses" would be to put extra (translucent) screens around your existing device screens (the physical screen can hide a keying pattern in a subliminal refresh rate) - so you look at your apple watch, and the glasses put a grid of 8 AR watch screens around the real one, so you can swipe inwards to switch the app thats active in the middle. Same for phone, and for Mac - palettes for your media apps being floating, hand addressable windows outside you screen, for example. It wouldn't work for a colour picker which needs opacity - but it could be virtually attached to your phone, so when you want to pick a colour, you look at your phone, whose screen becomes a colour panel, and the glasses bloom controls off to the side for interaction.
It's going to be a thing to increase the value of your existing screen, not replace it </my-suspicion>.
On step at a time. Let’s see them release desktop computers on a regular schedule first.
I just returned here after a week of trying to get some stuff done. Came to a simple realization, I will have to update the PC with at least one 4090 or maybe a rtx 6000 ada. and mac are simply not an option at the moment for me in the desktop. On the go it’s fine. if apple release something powerful I’ll look into it. If it has something to offer it might be interesting. Until then….
Proof that Macworld has gotten more and more stupid and irrelevant over time. Full of puffed egos and empty of intellect, insight and talent. I officially cancelled them with this and block them in all my news sources moving forward. Irrelevant blather…sound and fury of coward, sold out, wannabe losers:
www.macworld.com
Does Apple even need to make a Mac Pro anymore?
Between the limited upgradability of Apple Silicon and the power of Mac Studio, a Mac Pro doesn't seem to have a place anymore.
www.macworld.com
On the face of it, the shift to an SoC architecture rules out expandable towers like the Mac Pro. Perhaps Apple will surprise us with an ingenious technical solution, and we'll all feel silly for having doubted them. But there have been absolutely no leaks to suggest this so far, and the only titbits we've heard suggest failed attempts at a 4-way Ultra.
I'm sure MacWorld have nothing against Apple releasing a new Mac Pro tower. It wouldn't cost them anything, and it's another product line to review and speculate about. They're just speculating on the economics of producing it, and whether there's a big enough market to support it.
Admittedly, MW are also making the case that a machine like this doesn't need to exist, which isn't true. MacOS as a platform will be diminished if there's no workstation (i.e. large GPU) option. And in ways that even laptop users won't like ultimately.
I'm sure MacWorld have nothing against Apple releasing a new Mac Pro tower. It wouldn't cost them anything, and it's another product line to review and speculate about. They're just speculating on the economics of producing it, and whether there's a big enough market to support it.
Admittedly, MW are also making the case that a machine like this doesn't need to exist, which isn't true. MacOS as a platform will be diminished if there's no workstation (i.e. large GPU) option. And in ways that even laptop users won't like ultimately.
Last edited:
well more then 1 disk is an thing that pros need and it's nice to have that in one case.Proof that Macworld has gotten more and more stupid and irrelevant over time. Full of puffed egos and empty of intellect, insight and talent. I officially cancelled them with this and block them in all my news sources moving forward. Irrelevant blather…sound and fury of coward, sold out, wannabe losers:
Does Apple even need to make a Mac Pro anymore?
Between the limited upgradability of Apple Silicon and the power of Mac Studio, a Mac Pro doesn't seem to have a place anymore.
Nice to have USB dongles for some pro app in side the case is nice as well.
Room for non apple storage in side the case as apple mark on it is insane + pci-e 4.0 / 5.0 is just as fast or faster.
Other pci-e IO cards
Network cards other then copper 10 Gigabit Ethernet.
Well the mighty suck ***es of Macworld have decreed since they do not need or want it, you shouldn't either. And they certainly know more about what your job entails and what you need than you do. /sarcasm
It very likely won't. TBv5 isn't coming any time soon. By the time TBv5 (and x4 PCI-e v4 backhaul ) comes x4 PCI-e v5 SSDs would be on the market.
At best, the problem won't get worse if TBv5 eventually arrives. Unless the sole primary target is 'years old stuff'.
Thunderbolt is pragmatically even more Intel bound than it was before the baseline protocols got shipped out to USB-IF standards body. It is mainly about getting around the 'holes' the USB-IF standards puts in by declaring lots of stuff as 'optional'.
And switching to USB-IF standards body for the general protocol more glacier progress there also. Folks are just getting a wide variety of plain USB4 out the door. ( a standard 'passed' in 2019 ).
Apple also likely isn't looking to do a lot of PCI-e v4 on most of the SoC either ( for Perf/Watt reasons). [ For the Mac Pro the best hope is that large PCI-e v4 (or higher) provisioning is shoveled off to another die that has a looser Perf/Watt target than the CPU/GPU/NPU etc. ] Whatever slow poke pace the general USB-IF market adapts , Apple is more likely going to track that than to try to 'outpace' Intel.
There was a notion that Intel was going to weave TBv5 into Gen14 (Meteor Lake). That doesn't seem to be happening at all from the leaked info this far ( which is likely pretty complete since Gen14 is going into mass production. ).
Neither Intel , USB-IF , or the PCI-e standards body view Thunderbolt as a 'replacement' for PCI-e. So that really is not one of the primary Thunderbolt market objectives. There is a subset that it targets, but it isn't out to consume the whole thing. The more CXL 2 (and higher) gets entangled in PCI-e v5 and up
'value proposition' adoption that is even all the more true.
TBv5 also has a bigger role in bringing full DisplayPort v2 (and up ) provisioning to the TB networks also. (not the PCI-e part. )
I mean, they kind of have a point.
Let's imagine a fictional world where Ferarri, a decade ago, decided that they would make tons of money by focusing on super fuel-efficient daily-driver vehicles for the masses. So they made a few final supercars, and then went all-in on small engines with good performance for family cars, and spent the last 10 years transitioning completely to the small-engine family car business model. If Ferrari's supercar fans demanded that they keep making $250,000 high performance vehicles, would it be practical (or profitable) for Ferrari to spend time and money to try to adapt their fuel-efficient family car engine into a new supercar?
The reality is that iPhone was too successful for Apple, and it was the catalyst that pivoted them to becoming a mobile-first computing company.
Their points are atop their heads.
Great example. They would sell ZERO. The reason people would buy the boring cars is because of the halo cars. See Porsche. Try again.