Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
I’d be shocked if this wasn’t rolled into their 2030 carbon neutral roadmap as a KEY part of it. Apples data centers (currently including Azure and AWS) consume enormous amounts of energy
Serverside Swift is a thing. I would like to see it get a lot more support. I would also love to see Apple release a stripped down headless version of MacOS.
 
  • Love
Reactions: jonblatho
I think it's too soon to say. 8GB of unified memory is not the same as an 8GB ram dim. It could be the equivalent of 16 GB as we know it in an intel machine. IOS devices perform great and they have the most sucktastic ram numbers out there.

Some of these earlier reviews (Mac World had a pretty good one) are suggesting that it feels like its way but they don't have supporting data yet. In the Power PC days, less hz in the processor and less ram was faster and more efficient than Intel Counterparts.

I'm impatiently awaiting UPS to arrive.... and I have a massive and ugly Excel file that 16GB of ram on an i7 machine chokes on. If the M1 machine with
I want to get a Mac Mini but kind of waffling on the ram. I have an instinctual issue with spending $200 for an extra 8gb of ram. I recently bought 32gb of 3200mhz ram for my windows machine for under $130.

I realize that hardware profits make the Apple train run so the unified ram may make the difference for my fairly light use or maybe I'll over ride my instincts and spend the extra $200 anyway.
 
My guess is that native Apple Silicon Docker images with BSD and Linux are or will be available for M1 Macs soon. I'm waiting on Apple Silicon Docker images with Tensorflow to see how fast these new beasts can do ML training.
There is a tiny challenge with this right now. No Docker for M1, yet. It is coming, but it seems the complete rewrite for Apple hypervisor takes some time.

After that almost all Dockers will run on M1. Except...

TF will probably compile and run on M1, as it runs on any ARMv8 CPU. However, what you really would like to have is TF using CPUs, GPUs and Apple Neural Engine. That requires creating a lot of backend code possibly for APIs which are not documented.

I do not claim to understand too much of what might be involved, but Apple may not help here, as they want people to use their coreML instead. Also, it is not obvious whether ANE is any good in typical TF applications. (OTOH, Apple has used terms Tensorflow and M1 in the same sentence.)

Even the GPU road is a bumpy one, as there is no CUDA and Metal API seems to be ill-suited for ML tasks. But CPU, yes. Probably soon.
 
  • Like
Reactions: CJ Dorschel
I’m genuinely curious to know why they have not introduced a 16” version of this. Or let alone a timeline for an iMac version of this. I mean it certainly seems like a very viable alternative. Or maybe there is some catch. Or do they plan to just cause all of us to collectively **** our pants when they do bring out the new Pro-line M1s with double the performance over anything current at the time?
in addition to what others have said, the M1 appears to have some limitations in terms of I/O capabilities that would likely be deal-breakers for most buyers in the market for a 16” MBP. M1 machines are limited to 16GB of RAM, two TB ports, and can only support one external monitor (for MacBooks at least). I’ve no doubt Apple has an M1X or M2 in the works that will address these deficiencies and likely include even beefier graphics.
 
Well that is really excellent. I hope that Adobe will completely rewrite code for Lightroom Classic so it utilizes all CPU and GPU cores available. That would make large exports of RAWs to JPEG like a breeze.
We just tested a 50 RAW 42MP Photo export test using the M1 MacBook Air with 8GB of memory. Lightroom classic running under Rosetta. Finished faster than the $1800 13" MacBook Pro 2020 with 16GB of RAM. UNDER ROSETTA! (no optimizations at all)
 
Honestly, I think they were so confident (rightfully so) with their offering that they’re opting for as close to a genuine word-of-mouth marketing strategy as a massive corporation can do.

WWDC is going to be *ABSURD* this year....
They’re probably avoiding direct comparison with Intel Mac models for the time being because, well, they’re still for sale.
 
in addition to what others have said, the M1 appears to have some limitations in terms of I/O capabilities that would likely be deal-breakers for most buyers in the market for a 16” MBP. M1 machines are limited to 16GB of RAM, two TB ports, and can only support one external monitor (for MacBooks at least). I’ve no doubt Apple has an M1X or M2 in the works that will address these deficiencies and likely include even beefier graphics.
Is that an actual limitation though or a design decision for product separation and cost efficiency? Apple may well have done this on purpose because a) their research shows the vast majority of customers at this price point don’t use >16Gb RAM on a base level CPU b) the majority of customers at this price point don’t use significant I/O. So, there’s money to be saved here which can be passed onto the customer.

Those customers who have a need for more RAM, generally need a beefier processor and more I/O as well, and generally pay more for those machines with greater thermals. I think now that Apple control everything, we’ll see more of this type of product separation and efficiency savings.

Edit: “And though the M1 machines currently occupy the lower-end of Apple's Mac spectrum, the three devices that were refreshed with Apple Silicon made up about 91% of Mac shipments in the past twelve months, Morgan Stanley estimates.”
 
Last edited:
After 10 minutes of 100% CPU usage. Not many are going to need that in an AIR.
In my work, a complete recompile takes three minutes at the minute, which would go down to 1 minute 20 seconds. Another task I do rarely is 80% multithreaded and 20% single threaded, which unfortunately means the same amount of time is spent running multithreaded and singlethreaded. That part runs in 25 minutes, which would go down to 10 minutes, but for 5 minutes only one CPU would be running. So in neither case would I expect much heat throttling.
 
  • Like
Reactions: calstanford
Is that an actual limitation though or a design decision for product separation and cost efficiency? Apple may well have done this on purpose because a) their research shows the vast majority of customers at this price point don’t use >16Gb RAM on a base level CPU b) the majority of customers at this price point don’t use significant I/O. So, there’s money to be saved here which can be passed onto the customer.

Those customers who have a need for more RAM, generally need a beefier processor and more I/O as well, and generally pay more for those machines with greater thermals. I think now that Apple control everything, we’ll see more of this type of product separation and efficiency savings.
The M1 processor is the _low end_ processor. It has replaced low end and mid level Intel CPUs. A mid level CPU will be out soon.
 
Well that is really excellent. I hope that Adobe will completely rewrite code for Lightroom Classic so it utilizes all CPU and GPU cores available. That would make large exports of RAWs to JPEG like a breeze.
What makes you think there will be a "complete rewrite"? 99.9% of the code won't be touched.
 
  • Like
Reactions: RalfTheDog
I remember some members posting Apple was hiding something by only using generic graphs for performance measures during the Nov. presentation.
I think they were in a little bit of a tough spot because they are still selling those high end Intel macs at full price. Might not help their sales to show them getting embarrassed by sub-$1,000 macs. Apple definitely has to walk a line for a little while.
 
I wonder how many Mac Pros Apple are going to sell before they move to Apple Silicon.
Businesses and creation professionals will still need that power. They will still be buying.

When the AS versions of the heavy iron are ready, they'll judge and review that then.
 
Too bad the new laptops can't run Word for Mac 2011 and Excel for Mac 2011!

Seriously... I’m not kidding.

The last time I played around with a recent version of Word or Pages, or even the two leading Open Office suites, I was underwhelmed. Key features, extensive customizing, etc. were missing. Screwed up my writing and data analysis workflows.

Clearly, I’m going to have to give them another whirl — or, remain content with older hardware that still works amazingly well, and speedily!
 
Is 8GB going to be a problem if I get one?
I'm operating on a 2018 MBP 13 with only 8GB and its serviceable, but I do wish I'd gotten 16GB. I would say 16 should be the baseline for anyone who wants to use this machine for the next 5 years, and users should consider 32 if you're a power user or use any pro-level apps.
 
The M1 processor is the _low end_ processor. It has replaced low end and mid level Intel CPUs. A mid level CPU will be out soon.
I am very curious about the higher end processors. More cores? What configuration? 4 ice + 8 fire? GPU with more cores? More cache?

Higher clock frequency is unlikely, as it already is very high. The memory bus is already very fast.

I guess Apple can still do something more — but what? Keeping thermal limitations in mind.
 
Is that an actual limitation though or a design decision for product separation and cost efficiency? Apple may well have done this on purpose because a) their research shows the vast majority of customers at this price point don’t use >16Gb RAM on a base level CPU b) the majority of customers at this price point don’t use significant I/O. So, there’s money to be saved here which can be passed onto the customer.

Those customers who have a need for more RAM, generally need a beefier processor and more I/O as well, and generally pay more for those machines with greater thermals. I think now that Apple control everything, we’ll see more of this type of product separation and efficiency savings.
The fact that Apple replaced just the lower end 2-port MPB and not the 4-port model as well implies to me that it might be a hardware constraint, especially since the M1 can clean the clock of any U-series 10th gen Intel chip in the MBP 13 right now. It just doesn't make sense that Apple wouldn't have updated the entire 13" line with more ports or RAM options (which would have also benefited the Mini) if the chips were capable of it. The M1 itself seems like it might simply be a rebranded A14X in disguise so maybe the limitations make sense given that an iPad doesn't need that level of I/O.
 
I'm operating on a 2018 MBP 13 with only 8GB and its serviceable, but I do wish I'd gotten 16GB. I would say 16 should be the baseline for anyone who wants to use this machine for the next 5 years, and users should consider 32 if you're a power user or use any pro-level apps.
I wouldn’t buy a computer with less than 16 GB RAM, but...

My previous MBP had 16 GiB, and the current one 32 GiB. I was almost certain the difference would be huge, but to be honest, I have not noticed any difference with my workloads. I handle some rather large datasets and occasionally edit videos in 4k.

I think part of the story is the very fast SSD we have nowadays. Of course, fast SSD is 2 GB/s, whereas fast RAM moves the decimal point by at least one digit. Still, unless your algorithms require random access to the data, many things including video editing are acceptably fast even with some of the data on the disk.

I have never had a computer with too much RAM. However, 16 GiB may actually be a bit more than it used to be a few years back. (YMMV, of course; a billion point FFT screams for 32 GiB RAM.)
 
  • Like
Reactions: calstanford
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.