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.
Apple Had M1 Mac Pro Ready to Ship Months Ago, Mac Mini Redesign Unlikely
- Thread starter MacRumors
- Start date
- Sort by reaction score
No one would touch Apple for large data set science apps. That space is EPYC now. Same with Advanced ML/AI.
Render farms are leveraging both Nvidia/AMD for various areas in GPGPUs. Nothing in that space Apple can touch.
Apple has very select workflows optimized with the M series which is how come they appear to do so well in those spaces. For general computing they fall behind considerably.
Pricing won't vary much as all versus current systems. Over 70% of the cost in building out the current system is in RAM and the GPGPUs.
Because customers who are currently buying the Studio Ultra are professionals who will instead just buy the Pro, because the difference in price will barely make a bump on their balance sheet for (presumably) a leap in performance + configurability and expandability. Mind you I don't see how that's a bad thing for Apple because it means more money in their pockets.
I can see a continuing market for a Studio Max: Mac Mini on steroids with a performance on par with the 16" MBP Max, but I am wondering who will be buying a Studio Ultra when the Pro is launched. Maybe I just don't understand that mid-ground market.
Well, I need the fastest computer Apple can provide, so if Mac Pro will be significantly faster than Studio Ultra, I will buy it instead of Studio.
I do not need PCI slots though, so if performance is comparable, I will go with Studio Ultra.
Lots of people would have chosen M1 Pro Mini instead of M1 Max Studio. Cheaper and it would have been good enough
Considering the price bracket we will expect it to fill, I'd anticipate it to be faster than a Studio Ultra. There's going to be a lot of 7,1 Pro users laughing at it if it isn't.
Yes, of course. The question is “faster how?”. I’m using Logic Pro X with a lot of virtual instruments. If it’s only multi core advantage, I’ll probably go with Studio. If single core performance will be significantly higher, I’ll go with Mac Pro.
Even the M2 Air's single core score (1881) beats a Mac Studio Ultra's single core score (1754), so I would expect any Pro, based at the very least on some kind of "turbo'd" M2 architecture, to wipe the floor with it.
The main selling point of the Pro was power aaaaaaaand...upgradeability.
I am looking forward to seeing how will the market in need of the "Pro" react if we get locked down non upgradeable "Pro"
I am looking forward to seeing how will the market in need of the "Pro" react if we get locked down non upgradeable "Pro"
Simplest most basic case. Trying to run a 3D EM simulation with Lumerical that requires 1.5TB main RAM to proceed or no go due to the size of the structure and the grid resolution needed to get the discretized Maxwell’s Equations to converge. I can do that with a 2019 Mac Pro or even older PC. Zero chance of performing that in the most modern highest performing iMac for the foreseeable future. That’s just a simple RAM case. There is also plenty of hardware used in engineering applications that you need a PC’s expandability to utilize.
Looking forward to seeing how the new Mac Pro will be with Apple Silicon.
You seen this?
Predictions (I made on this site
that Apple will release their own GPUs.
Also, user expandable RAM.
that Apple will release their own GPUs.Also, user expandable RAM.
Specific high-end use cases only support my argument that workstations are bad investments for most people.
Me: Race cars are expensive and impractical for daily drivers.
You: Some people need them to race professionally.
The 2019 Mac Pro is the current shipping Mac Pro. Until Apple replaces it, if you need a mac workstation that supports 1.5TB of RAM and multiple GPUs, that’s the one to buy. That’s why they sell It. But, you buy it configured to do work you have to do now, not to hedge bets against some future unknown expansion need. Apple is a pretty bad example because their last expandable Mac Pro wasn’t updated for nearly 8 years and people that needed the upper limits of the machine had no choice but to maintain them for years. But the point is that most people did not need the upper limits. Most cheese grater Mac pros never had more than one GPU, 32GB of ram, and a single SATA HD. Even in the PC world it holds true where people buying big Xeon boxes with entry-level specs and unfilled slots hoping to expand later are just buying an expensive dust vacuum to sit under the desk.
Last edited:
Is this guy on the MacRumors payroll or something? His “predictions” seem to be all over the site now, and most of them have been wrong. Why would you keep promoting this guy?
I think the single core advantage will be simply based on the generation of the chip and when you buy. An M2-based Mac Pro will have better single-core performance than an M1-based Mac Studio, but an M2-based Mac Studio that is surely in the pipeline will probably be on par with the Mac Pro. Unless they clock the Mac Pro higher.
Gurman on the payroll? Kind of ... more than several years ago he used to work for Macrumors. One of the authors on front page 'news' articles. Technically, not on the payroll now, but lots of alumni cheerleading. (If Gurman gets paid megabucks from Bloomberg maybe someone else hires out the macrumors writers pool later. ) He collects a check from Bloomberg , but the real change over last year or so is that he started a weekly newsletter ( think the sign up is on Bloomberg site). The 'problem' with a weekly (has to go out every XXX day of the week newsletter is that leaks/news doesn't necessarily come weekly.
So instead of " Apple source showed/told me X, Y, Z " get filler content of "i think , i suspect , apple should ..." content.
MaxTech very similar. Lots of clickbait to drive ad traffic for his channel. ( more than a few front page macrumors articles are headlined and synopsis written to drive clicks and long running commentary. ). Again ad traffic.
They also said that was the last M1. So it couldn't be a 2xM1 'something' either. All Apple said was that something was coming later. A hefty set of folks took later to be "WWDC". That assumption had major problems given the other statement they made implying that it would be something "after" the M1 generation. Some hand waved that off by saying that Apple would slap another prefix in front of the "Mac Pro" SoC because it was going to be super special. That is highly likely very deep wishful thinking. There is probably going to be significantly high overlap of basic building blocks amount the Mac dies. The Mac Pro should have a different allocation of "stuff" but very good chance it is primarily the same 'stuff" allocated onto the dies of various shapes and sizes.
Once Mac Pro waiting on "M2" generation, it is probably sliding well past June.
Apple 'covered' the iMac with the Studio. They are also marking the Mini as 'complete' the transition even though had not replaced the Intel Mini. Still haven't.
The Mac Pro has a decent chance of getting a different "Ultra" class solution as its starting point. The M1 Ultra wouldn't be all that great in terms of PCI-e I/O provisioning (unless looking to so a relatively weak 1 slot wonder. )
There isn't a good high performance way to double up the M1 Ultra. The UltraFusion there only tightly bounds two dies and then it is spent. And the chip packaging tech being used is at the recticle limit so would have to use something substantially different, larger, and more expensive.
The rumors have been the Apple has been doing a "ground up" refresh of the Mini chassis.
Apple's Next Mac Mini Refresh: Everything We Know
Apple is working on updated versions of the Mac mini, and at least one of the Mac mini machines in the works is likely to replace the Intel version...
www.macrumors.com
Going through Industrial design is a chokepoint. The supply chain hiccups are another. Similar reason why the MBP 13" has exactly the same other stuff an only got a minor logicboard tweak for the M2. If Apple is busy painting themselves into a corner on the Mini by thinning it out too much then that takes extra time.
Gurman "expects" doesn't necessarily mean that is what Apple is doing. [ Or what Apple wanted to do , but are now course correcting back to classic chassis after painting themselves into a corner. ]
If the case is the same, then probably limited on supply of M2 and waiting for the laptop sales to get to the initial demand surge to get to more quantity. If the Mini and iPad Pro launch in the Fall that would be follow on set of products to soak up 'plain' M2 chips. M2 seems to share the same process fab tech as the A14 and that too is ramping this summer for the non-Pro iPhone 14. Again limited supply issues.
The MBA and MBP 13" are the best selling Macs. There is little good reason to slow down their availability this summer so that could swap sales with Mini's.
Exactly. Right now there‘s a big gap between the M1 Mini and M1 Max Studio. An M2 Pro Mini would fill that gap nicely.
Not to mention on the really big dataset end of things you're talking clusters, which Apple was never a huge player in, was never a player at all in the HPC world at all, and definitely does not have the software stack to support it our right now even if they came out with new 1/2/4u boxes and appropriate LOM/interconnect support/etc