I ask this because when M1 Mac’s came I felt we were entering a new era of portable PCs: fast, lightweight and with a long awaited good battery life.
I just saw the announcement of the Snapdragon X Plus, which is looking like a response to the M chips, and I am seeing a lot of buzz around it, so I ask: what is so special about it?
In: Technology
The change from x86 to arm, the unified architecture, etc. allowed M1 to have way better performance per watt than previous desktop class cpus. This means way better battery life, thermals, etc. while retaining good performance in a small and lightweight package.
This doesn’t need to be exclusive to the M1, of course – other chipmakers are now doing similar designs, and will probably see similar results. M1 was revolutionary because it was the first to do it on a wide scale for computers (phones have been using a similar architecture for a long time).
The M1 chip allowed Apple to optimize it for their hardware and software ecosystem. Unlike traditional processors where the CPU, GPU, and RAM are separate components, the M1 integrates all these elements into a single chip. This unified memory architecture allows for faster communication between the CPU, GPU, and memory, reducing latency and improving performance. By having high-efficiency cores alongside high-performance cores, it could better balance power consumption and performance based on workload demands compared to the competition at the time. While it can’t compete with dedicated graphics chips, the M1 chip integrates a custom GPU designed by Apple, which delivers respectable graphics performance while also being energy efficient.
Imagine a powerful gaming PC as a heavy V8 muscle car that goes 0-60 in 2.8 seconds using brute force power and gets 16mpg and the M1 as a Mclaren turbo-charged V6 that is really dialed in and goes 0-60 in 3.2 seconds, but gets 30mpg. If you need a relatively powerful setup (not the most powerful) and long battery life, nobody can touch Apple with their custom silicon.
It is less revolutionary than evolutionary, the M1 is a RISC (reduced instruction set processor) using the ARM instruction set. Before the intel Macs I think motorola created the mac chip and it was also RISC like. So it isn’t like anyone reinvented the wheel here.
There is a misnomer another poster said, that the mac fused the graphics with the chip but graphics ‘on die’ have been on Intel chips for years. What Apple was able to do was share high speed memory between components that used to have discrete memory, and that greatly improved efficiency. As far as the other benefits of M1, like better heat and power requirements, this has been a feature of RISC based chips for a long time. They just were the best ones to break the hold of Intel/AMD on the market.
The M chips aren’t actually _that_ revolutionary, and the M1 were tested in iPads and then in iPad logic boards crammed into Mac Mini cases before releasing to market.
Apple has been rolling its own silicon for the phones and tablets since 2010’s iPhone 4, and that at the time of M1’s launch to market represented over a decade of design experience within the arm64 architecture. Now, what are key design features that are handy in a computer but life-critical in a smartphone? Efficient performance. You only have a few watts to play with (in the case of a phone contemporary to the M1, the A14 Bionic SoC has a TDP of 6 W), so you’re going to make the best of every single milliwatt and energy-saving trick in the book while still retaining class-leading performance. iPhones consistently bench high compared to Droids of the same year. When you make the jump to a desktop and all of a sudden you have up to ten times as much power to play with (10W for M1 passive cooling, double that for M1 active cooling, 30W for M1 Pro, M1 Max about the same, and 60W TDP for the M1 Ultra), you can work ten times as hard. Or you can work as hard, but extremely efficiently, because you now have what’s effectively a phone processor doing the job of a desktop processor just as well.
The revolutionary part was attaching a keyboard and mouse to an ARM chip and making lots of software available at launch, either native arm64 or x64 running in a virtual wrapper. Microsoft sort of missed the boat on that latter part on its ARM powered tablets.
TSMC introduced a new 5nm fabrication process in 2020 and Apple got first crack at it. Because of the quantities Apple was purchasing, they essentially monopolized the manufacturing line forcing everyone else onto older less efficient 7nm and 10nm processes until TSMC could build up 5nm capacity. The other companies that could get 5nm parts produced were mainly interested in the smartphone and GPU markets and not initially looking to produce CPUs.
To make a very long story short: M1 took out all the old inefficiencies of X86. Apple can do this because they’ve never cared about legacy systems and can tell their developers to figure it out or get lost. Intel has such a huge install base they can’t make giant changes like that. I mean they could, but a giant mob of developers and users would riot at their HQ.
Latest Answers