Taiwan Semiconductor Manufacturing Co., Apple’s partner for iPhone and iPad processors, will build the new Mac chips, said the people, who asked not to be identified discussing private product plans. The components will be based on a 5-nanometer production technique, the same size Apple will use in the next iPhones and iPad Pros, one of the people said. An Apple spokesman declined to comment, as did Intel and TSMC. Current mobile device chips from Apple have multiple processing units, or cores, that handle different types of tasks. The latest iPad Pro has four cores for performance-intensive workloads and another four to handle low-power tasks to preserve battery life.
The first Mac processors will have eight high-performance cores, codenamed Firestorm, and at least four energy-efficient cores, known internally as Icestorm. Apple is exploring Mac processors with more than 12 cores for further in the future, the people said.
In some Macs, Apple’s designs will double or quadruple the number of cores that Intel provides. The current entry-level MacBook Air has two cores, for example.
Apple will release the first Mac with Apple silicon at the end of this year, and it expects the transition to take two years.
Rosetta is a translation process that allows users to run apps that contain
x86_64 instructions on Apple silicon. Rosetta is meant to ease the transition to Apple silicon, giving you time to create a universal binary for your app. It is not a substitute for creating a native version of your app.
Rosetta translates all
x86_64 instructions, but it doesn’t support the execution of some newer instruction sets and processor features, such as AVX, AVX2, and AVX512 vector instructions.
looks like my arm architecture builds of vcvrack might get some unexpected new importance in the future … but i doubt that macos 11 will still support opengl … but still a while to go until all this gets reality …
the good news for vcv is that these instructions are rare (if even used) in most plugins, and the sse2neon header supports most of what is already used.
as of the pre-release notes, it still does, but those are “pre-release” so could likely change. hard to guess until someone (Andrew?) gets the new dev hardware in hand (it’s $500 for access, and is essentially a lease of the hardware - maybe a new fundraiser?)
I’m not too worried about it. Apple is losing the pro audio market fast, so people will almost completely switch to Windows machines by the time today’s Mac computers are obsolete. If their processor/OS architecture becomes the norm in 10 years, I’ll officially support it. But only if Apple is friendly to developers, which they’re not so their direction will have to change.
Reusing that name makes their game plan pretty obvious. The last time Apple called something Rosetta, around 2006, it was supported for like what, 4 years?
I have a Mac Pro from circa 2006 gathering dust - great machine that still works well enough. I never boot it up anymore, too much nostalgia for the glory days of 10.6… I prefer booting up my first-generation PPC mac mini I hacked to boot in OS 9 and bang out a beat in ReBirth.
Computer tech should just stop making progress and chill out for a few decades imho. Would be better for everyone.
VCV Rack excluded
Maybe if it stopped making incremental progress on technologies that date from the 70s and 80s (ie windows, macos, linux etc) and maybe caught up with developments from the 90s (BeOS, Taligent etc) we could all be living in the computer dream world they promised us!
I am very bitter.
Does anyone know how much apple IP is in these chips vs other people’s? Obviously Apple didn’t invent the ARM CPU, just like they didn’t invent the Power PC. I’m sure Apple’s contributed something to this, but what?
Well there’s a depressing thought. I know Windows inside out. I know my Mac pretty well. Never a question in my mind which one I’d choose (money aside). Mac is hands down better on all accounts in my book.
arm64 is an apple extension (not to be confused with aarch64), the power management, the chip design (apple does not use the arm reference designs), the memory controller, and apple’s neuro-chip are all additions that apple has made. these changes are coupled with 2 types of cpu cores.
add to this their in-house designed gpu’s, which are extremely nice, especially compared to intel’s.
remember, apple purchased and integrated multiple cpu design companies over the last 10 years and has a top-notch team.
samsung, who also develops their own cpu’s and do not use the arm reference designs, is a fairly close second, but benchmarks still have apple out ahead: https://nanoreview.net/en/soc-compare/samsung-exynos-990-vs-apple-a12x-bionic
Interesting, thanks for sharing.
@Vortico couldn’t Rack be built on the new Macs the same way it is on the old ones? Obviously any closed source module developer would need to get a New Mac (or cross-compile). But the strength of open source software is portability.
This is assuming Apple doesn’t put other obstacles in the way of users building their own software.
Osborne effect coming…
(I think my next DEV machine will be a macmini m1)
Sorry man, but this thing will become impossible for you to ignore. The benchmarks are unbelievable.
That part is still relevant however. But i do agree, these new benchmarks are impressive, and i recall someone mentioning in another thread here that this might cause a Osborne effect hehe
I’m someone who has worked on all sorts of computers, starting with 3270 terminals connected to an IBM mainframe. I’ve done serious development work on mainframes, DEC computers running Vax/VMS, CP/M80, MS-DOS, Windows, OS X, and Linux. The bottom line on all that work is a computer is a computer, you figure out the platform quirks and get on with it.
Apple is hostile to developers in so many ways. @Vortico is developing a closed source VST plugin for Rack. From what I hear Apple will require digital signing of any software, make him jump through hoops to get into the App store, require him to buy a new computer to do development for M1 Macs, etc etc.
Apple’s argument is that it’s all to provide a positive user experience. But it seems motivated more by a desire to make the platform their “walled garden.” They will - grudgingly - allow users to build & run their own software, but I suspect that will grow more & more difficult.
They want control of their platform. Linux & Windows are wide open. That can have a downside: it’s easy to trick users into running malicious software. But the upside is greater: users control their platform, most of all on Linux, where a dedicated user can change the OS itself on a fundamental level if they want.
Apple, more and more, wants to have their expensive dongles in everyone’s pocket. The desktop market is an afterthought at this point. And their platforms do offer a lot in terms of ease of use, but Windows isn’t harder to use, just different. Linux could use some good ‘productizing’ but no one has stepped up to build the infrastructure to make Linux completely turnkey. It’s not easy to get a proper pro audio workstation going on Linux. I know because I try to do that periodically and it’s maddening the amout of low level BS you need to go through to get Jack & software configured and work.
Apple will less and less be a serious platform for technical users. Research (Academic & Private) has moved more and more to Linux, because it’s a scalable platform that’s easy to develop for. Windows will end up being where creatives end up, because the Apple platform will just keep getting more expensive and harder to work with.
It’s a shame because Apple’s marketing to creatives over the last 30 years has been a stunning success. Such a shame they’ve chosen to make developing creative software so difficult. You can literally buy a computer with no OS, put Linux on it, and have access to every serious tool for software development at your fingertips. There’s no cost except for the box & monitor etc.
With Apple you pay an exorbitant amount for the computer, then pay again to sign up as a developer, and pay again when you give them a cut of the price of your software when it’s sold in their app store. Plus they require you to learn their GUI development tools, and ignore international standards, both official and de facto.
just a little extra note as many people seem to very impressed by the benchmarks: i think so far nobody external really tested this new hardware for a while in real world scenarios - if the new apple silicon will stand out that much then still, then it might be the time to get excited, but for now this all sounds just like clever marketing to me (just check their wording carefully - things like “up to x times” …) - in the end even apple is bound to the laws of physics and they might get a small edge over other arm cpus if they are good or lucky but they will not be able to build anything which is by the order of magnitude better than the others, as they are trying to sell it …
best wishes - hexdump
AFAIK a lot of the buzz started around the performance of the dev kit mini with the iPad chip. We might not have real world test results from the M1, but having the dev kit as a reference point helps the case for expecting the M1 to have impressive performance in practice.
Fwiw here’s one softsynth dev who has been working with the dev kit talking about his expectations for the M1 + arm mac chips: https://llllllll.co/t/which-macbook-should-i-buy/9601/431?