Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
If the M2 is faster than the M1 Pro then it probably wont be on any entry level models. But its confusing to have M1 chips that are faster than M2. I think Apple does not want to go back to the days of comparing megahertz to see what chip is faster. A clear distinction between pro and consumer chips would be easier to tell apart. Something like dual core on prosumer models, quad chips on the pros and single m2 processors on the low end would be nice. Having clear speed differences between models is preferable too. That way when your spending $1,000 or $2,000 more than the low end, you feel that difference in everything, not just the hardest tasks.
 
If the M2 is faster than the M1 Pro then it probably wont be on any entry level models. But its confusing to have M1 chips that are faster than M2. I think Apple does not want to go back to the days of comparing megahertz to see what chip is faster. A clear distinction between pro and consumer chips would be easier to tell apart. Something like dual core on prosumer models, quad chips on the pros and single m2 processors on the low end would be nice. Having clear speed differences between models is preferable too. That way when your spending $1,000 or $2,000 more than the low end, you feel that difference in everything, not just the hardest tasks.
Apple made a decision to 'de-risk' their own silicon roll out by starting with the 'consumer' models. This gave developers plenty of time to support the new architecture so that by the time the 'pro' models were released, a sufficient quantity of the application library was ported to Arm64.

The consequence of this is that the consumer models are leading the pro models now. Given the incredible performance of M1, it doesn't make sense to me for Apple to release an M2 right now. They should continue to move their entire line of computers to Apple Silicon first. This means they should push out an iMac Pro this Spring and then ideally a Mac Pro in the summer/autumn. They could then release updated consumer models early next year either using an M1X type name to avoid positioning the chip ahead of M1 Pro/Max from a marketing perspective, or release updated consumer and pro models at similar times with M2, M2 Pro, M2 Max.
 
Gurman is a clown
You seem really upset about that. But then again, when a random Internet stranger calls somebody else a ‘clown’ anonymously, I wouldn’t take anything that person says seriously.

I would rank Gurman in the top three analysts, although he’s not entirely always accurate, neither is Kuo, who has the highest reputation, given his insider knowledge with Apples suppliers. Any time when you have somebody that makes predictions about a future product for a company as major Apple is, I don’t think others expect complete accuracy, more of the possible idea of what to look forward to in the future, which doesn’t include delays, hardware alterations, what does and doesn’t release, etc.

But one thing all these analysts have in common, is rehashing the same rumors to try to stay relevant and competitive. But I would still rather read what he has to say, than not.
 
Last edited:
You really give people that little credit? That someone wouldn't be able to understand the difference between an m2 and an m1pro?

People sure did fine with the A12x and a13
For the overall general public? No. I don’t. I think Apple’s lineup is already confusing enough. Air, pro, max, mini… look at the iPads, it’s a mess. And people care a lot more about a processor in a Mac than an iOS device. Again, I don’t think they should come out with a new processor every year like they do with iPads and iPhones.
 
For the overall general public? No. I don’t. I think Apple’s lineup is already confusing enough. Air, pro, max, mini… look at the iPads, it’s a mess. And people care a lot more about a processor in a Mac than an iOS device. Again, I don’t think they should come out with a new processor every year like they do with iPads and iPhones.

So true
As the longtime "tech" person in my extended family, I get questions all the time about iPad purchases (and all the Apple stuff really).

Nobody has any clue why they should choose an iPad, iPad Air, iPad Pro, or which size would be best...or should they go refurbished? (like what changed? Should they care?). What about last years model?

It's a big hot mess.

Tim loves a big messy SKUfest product lineup.

Say what folks want about Tim's Apple -- a big drawback to his approach to running things is the fragmented crap show the lineups have all turned into.
 
Waiting for a new MacBook Air with smaller bezels/bigger screen. Hopefully no notch, and hopefully no plastic/fake/cheap feeling to it like the new MacBook Pros… or if they go for the cheap build feeling with the new MBA too, then keep it below 1000$
 
Last edited:
I love how Gurman can squeeze articles and newsletters out of the same rumors over and over again. Tells you something about how thirsty we are for Apple rumors.
Because it’s a very secretive company that’s why. It’s like a tease!
 
If the M2 is faster than the M1 Pro then it probably wont be on any entry level models.
It wont, once the cpu can only match the 10 core m1 pro but on the gpu side cannot come even close to the binned m1 pro with 14 gpu cores
M2 probably also dont even have 100gb/s bandwidth, compared with m1 pro 200
 
Last edited:
I have M1 MBP with 8GM of ram. Am seriously considering an M1/M2 Mini to use alongside it because having too many Adobe apps open eats up the ram. Not just the ram but the scratch disk too, I got the 256GB model. I should probably save for the M1X instead.
You really should’ve picked 16GB of RAM and 512GB storage on the get go. I picked 16GB 1TB M1 MBP and I still experience beachball and mid-to-high memory pressure fro time to time.
 
The M series of chips is referred to as a "System on a Chip", or SOC for short. If the "system" is an upgrade, why do you care about a speed gain? It sounds like you haven't gotten off of the Intel "MHz/GHz train", and that's a train that left the station what, 15 years ago?

You should reconsider whether or not you're putting too much emphasis on speed ratings in your thinking and/or purchase decisions.
I understood "speed gain" to mean an increase in performance, rather than anything directly related to clock speed etc.

It typically means higher benchmark scores, leading to faster completion of tasks - hence the "speed gain".
 
The PPC architecture and ISA wasn’t massively scalable at the time. That’s why it died. ARM is a little different as it’s just an ISA and the implementation is vendor specific. Apple have total design control here.

Don’t forget Intel dropped out the Pentium 4 at the time when it switched architecture to compensate and it was a disaster for a whole CPU generation. It wasn’t until after that debacle when the Intel Core line appeared that things started looking good again.

You learn from your past mistakes. In this case apple’s mistake was trusting vendor architectures. I suspect they will stop trusting vendor ISAs in future as well and will migrate from ARM to something in house.
I don't see ARM as a "vendor ISA", at least not in the same way as x86_64 as implemented by Intel & AMD; It's a "paper specification" from ARM for which Apple has a perpetual licence and is already adding their own extensions. Apple has already migrated to something "in house" because they control the entire SoC design.

Unless the ARM ISA was found to have serious limitations, Apple is not constrained by it, and can extend it to their hearts content.

I would argue that Apple has far more to lose by moving to a completely proprietary ISA - they would immediately lose 100% compatibility with all ARM64 software and all the migration work to Apple Silicon that has already happened. They would be *hugely* unpopular and lose support of software developers. Unless of course, they have a new version of Rosetta that translates ARM to Apple-RISC....

I really can't see this happening...
 
I have a sneaking suspicion the Mac Pro will stay intel for a few more years.
They might keep selling the 2019 Intel Mac Pro for a year or two, but I would be astonished if Apple didn't pull out all the stops to release an Apple Silicon Mac Pro before the end of the year.

Failing to meet their own timeline for fully transitioning to Apple Silicon would have a terrible impact on their credibility (and stock price).
 
i think it will. i think the BIG mac pro will forever be intel because apple silicon cannot support that form factor as far as we know. those pcie slots will be useless. i think the plan is that they make a smaller mac pro or bigger mac mini that supports the strongest apple chip, and leave the desktop computer case sized mac pro for intel. there were rumors that they was gonna update the intel chip.
Why can't Apple support the Mac Pro form factor?

Apple Silicon already supports PCIe - it's how it supports Thunderbolt. And Linux drivers for PCIe are currently being written for M1 Silicon (https://www.phoronix.com/scan.php?page=news_item&px=Linux-5.16-PCI). Better tell all those devs not to bother...

The ARM designs already support PCIe so there is nothing inherent in ARM or Apple Silicon that prevent its use - how do you think all those ARM servers work? e.g. https://www.anandtech.com/show/1594...ytes-2u-with-80-arm-n1-cores-pcie-40-and-ccix

Your user name is apt...
 
SOftware macOS meeting:
* Updates in GarageBand, Photos, iMessage, Mail, Contact, Calendar ...
* Deprecation of Pages, Numbers and Keynote in favour of LibreOffice

:)
 
Was hoping they would update the iPad pro. Would love touch Id under the home button. Any idea when they might refresh the pro lineup?
 
At what point is the m1max delayed so much we move on to m2max?

M1 came out November 2020, and the same chip with more cores still hasn’t been released almost 18 months later..
The feeling is that Apple will release a new M chip every two years. Which is good, because it also mean devices will hold their value for longer, and there could be less waste to pollute this planet with.
 
With these analysts and leakers, I wouldn't trust a single word of their timing and launch windows, or renders of what they think the devices will look like.
Apple frankly embarrassed them ALL over the last 2 years as they repeatedly got things wrong.

I trust the specs and the actual devices they claim we will get though.
 
Please let this be the announcement for the 27" or larger M1-series iMacs. I am literally like Fry holding out my stack yelling "SHUT UP AND TAKE MY MONEY" here. If I can buy that in the first half of this year, and work issues me an M1 Macbook Pro like I suspect they will, my life will be complete.
If you're waiting for your life to be complete, just know that anything material will not do it.
 
I love how Gurman can squeeze articles and newsletters out of the same rumors over and over again. Tells you something about how thirsty we are for Apple rumors.
I don't get why people are upset about this. He writes a weekly newsletter—which is always an insightful read—but it's MacRumors who decides to keep posting articles as if it's brand new information.
 
Why can't Apple support the Mac Pro form factor?

Apple Silicon already supports PCIe - it's how it supports Thunderbolt. And Linux drivers for PCIe are currently being written for M1 Silicon (https://www.phoronix.com/scan.php?page=news_item&px=Linux-5.16-PCI). Better tell all those devs not to bother...

The ARM designs already support PCIe so there is nothing inherent in ARM or Apple Silicon that prevent its use - how do you think all those ARM servers work? e.g. https://www.anandtech.com/show/1594...ytes-2u-with-80-arm-n1-cores-pcie-40-and-ccix

Your user name is apt...
It’s not about the PCIe but what you do with it.

90% of system configurations are 100% static and will never change once purchased. No one adds CPUs. No one adds RAM. No one adds GPUs. Occasionally people add storage and that’s about it. It makes sense at that point to optimise for cost and shipping efficiency which means smaller form factor devices which are cheaper with static configurations. On that basis it makes sense to push all those PCIe lanes out over Thunderbolt and/or internal NVMe slots and 10G ethernet and that’s about it.

If you need compute it’s far far cheaper to rent it than to buy it these days as well so the local processing is almost always fairly limited compared to a decade ago. For example if you’re doing content delivery, then transcode at the edge on the CDN is where it is. If you’re doing really big simulations, you’re going to rent a cluster for the simulation window not buy one to run locally. For other use cases such as video / media, what we have now is INSANELY POWERFUL. Thus the problem domain is constrained.

There is also the situation where kext loading is forbidden in macOS so only standard shipped drivers will be available. Random PCIe slot stuff isn‘t going to work. That’s not really a problem though as there are very few non-generic devices out there now. Most seem to be USB if they are with stupid device IDs (Don’t get me started on that)

Optimum system configuration for a first gen “Pro” device is probably:

* Mac Mini like form factor
* Up to 2x M1 Max SoC
* Up to 128Gb RAM
* Up to 2TB SSD
* Up to 2x NVMe slots (I bet this won’t happen)
* 4x thunderbolt ports
* 4x usb ports
* 1x 10G ethernet
* BT / WiFi 6
* Gigantic external power brick that gives the TB/USB ports PD capability.

They could do a larger one with 4x M1 Max SoCs.

Not sure what architecture. I suspect this scales without having to resort to NUMA (yuck)

Also that above isn’t going to require £700 wheels because you can just pick it up and steal it rather than wheel it out of the office door under cover of darkness ?
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.