I keep hearing about the new Macs with ARM processors, as opposed to x86. I don’t understand what either of those mean.
In: Technology
Processor architecture means the way the processor carries out instructions. x86 is designed to be able to do a bunch of specific complex instructions and ARM is designed to do a limited group of very simple instructions.
But a long time ago people realized that there’s no limit to what complex things we’ll want to add to the list of automatic complex instructions, and that every complex instructions added to the architecture slows down its ability to do ANY instruction. But you can also do any complex instruction by breaking it down into very simple instructions.
These things put together mean that the more complex things you want to do with a computer, the faster it will be able to do those things if you just build them out of many simple instructions rather than fewer complex instructions.
The problem is, computers were already mostly using x86 when people realized that, so for a couple decades we’ve been using the less efficient system just because it’s what everyone uses and the companies that make processors don’t want to have to change.
It’s like if you’re drawing. It’s easy to make straight lines, and triangles, and circles, and squares. And with those shapes, you can draw literally anything else. But if someone gave you a magic pencil that could automatically draw a face, drawing faces would be SO much easier. But if it can only draw the same face then you either have to settle for the fact that all the people you draw will always have the same face, or you’ll only be able to automatically draw one face and still have to do the rest by hand any time you draw a group of people. And it’ll be completely useless any time you draw a bridge. But if you have a pencil that automatically draws basic shapes, drawing ANYTHING becomes a lot easier.
You may think of ARM vs x86 as being similar to the difference between programming languages like C++ or Python but in this case, it is the hardware language rather than a software language.
The significance of this change is that Apple will jump ahead in the laptop market at a minimum and who knows what they will do with a desktop computer. ARM is known for having low power consumption with good performance and apple has been making the best mobile processors you can get for a while now by using ARM. Until this change, they have only used these mobile processors in their phones and tablets. Now they will bring this incredible efficiency to their laptops. You can expect a large boost in battery life and the devices will probably get even smaller as they may no longer need cooling.
There are a variety of differences between ARM processors and traditional x86 processors – if you’d like to get into the technical differences, there’s lots of articles that discuss it if you search the web.
At a high/simple level, ARM and x86 processors have different instruction sets. ARM processors focus on efficiency – if the processor is going to process some data, let’s process it as efficiently as possible from a power/electricity consumption perspective.
x86 processors focus on performance – if the processor is going to process some data, let’s process it as quickly as possible no matter the power/electricity consumption.
This makes ARM processors ideal for applications where power is limited, such as battery operated devices (IE. cellphones and laptops, or even high performance rigs where power consumption is extremely high and may need to be reduced as much as possible).
x86 processors are ideal for applications where power efficiency is less important, like a desktop computer.
ARM processors currently have the highest density of transistors along with better power efficiency. They’re not far from catching up with x86 CPUs in processing power, at which point we’ll probably have laptops with all-day battery life, or you’ll just ditch your laptop and use your iPad or iPhone for all your computing needs, connecting it to a mouse/keyboard/monitor at home.
Processors run on what are called ‘instruction sets’, basically the hardware is set up to perform calculations in a specific way.
x86 or x86-64 has been the standard in Desktops and Laptops for the last few decades and the x86 instruction set is built around versatility. The downside comes down to the old adage of “A Jack of all trades is a master of none”. An x86 processor can do almost anything you’d want it to do, but because it’s not specialized in any one thing, it’s not going to be particularly good or efficient at the tasks you ask it to do.
ARM based processors became popular mostly due to Smartphones. You don’t need a lot of versatility with Smartphones (at least you didn’t when they first became popular)…so rather than go with a processor that can do lots of things with okay performance, you could design a processor to do a few specific things particularly well.
The advantage of this with mobile devices is that when you design a processor to perform very specific tasks, it can do it a lot more efficiently, use less power to do it and basically be better at it. The downside, of course, is that you’re going to take a major performance hit when you ask it to do something it wasn’t specifically designed to do.
Just a different type of chip. A different set of instructions, with different capabilities and different timing.
They’re all pretty equivalent in the long run, but ARM is built to be made of lots of small, simple instructions (“Add this to this”) whereas x86 is more built towards more complex instructions (“Apply the mathematical function to this matrix of numbers”).
It’s like deciding what language they talk – you can talk “simple baby language very fast” or “adult language where one word has a lots of implications and so may take some time to comply with completely”.
In the end, both types of architecture have advantages and disadvantages and there’s no clear winner. They are both produced in their billions worldwide.
But with ARM in particular, just because of its history and its focus on how the chips are designed, the way it’s held the market is that their chips are powerful, simple and also extremely low power consumption for that job. x86 chips are one of the highest power-consuming chips.
ARM originally powered something called an Acorn machine, from the 80’s. They’ve been around that long. And they now power virtually every smartphone in the entire world (yes, including Apple).
The problem is that speaking ARM and speaking x86 are two very different things. Of course you can describe any particular task in either language, but everything that’s only been written in x86 needs to be “translated” to ARM to make it work properly, and vice-versa. Most modern commercial desktop operating systems are x86. But most phone and tablet operating systems are ARM. Hell, even Windows running on ARM is nothing new.
Fact is, that an ARM can do just as much, and everything that an x86 is capable of. It will likely use less power to do so (great for phones and tablets and laptops, not so important for a desktop computer). But most of your games / Windows applications are written in x86. That doesn’t mean they couldn’t be translated, but it’s not just as simple as “run them through Google Translate”. The creator of the program has to be willing to translate the program as only they have access to the “original text” (source code), and not everybody will bother.
So you know how computers have a little chip in them called a processor. If you send specific electrical signals to the pins of this chip, it will perform certain basic operations like arithmetic, comparing numbers, storing or retrieving numbers from memory etc. And this is how we can get computers to do things useful for us humans.
Of course you need to know what specific signals will perform which instructions, this is called the instruction set, it’s like the digital “language” that the processor speaks. ARM and x86 are two of the most popular instruction sets.
As for what’s special about ARM, well, nothing really. x86 is a type of Complex Instruction Set Computer (CISC), meaning it can handle complex operations with dedicated instructions. ARM is a type of Reduced Instruction Set Computer (RISC), meaning it has fewer instructions available, but can perform them more efficiently. Neither is “special”, they are just two different approaches to achieving the same goals. The line between the two gets fuzzier each generation, and each approach has trade-offs.
As an analogy: A truck and a motorbike are both popular means of transport, a motorbike will take less fuel to reach your destination faster, but good luck using it to move furniture.
Arm is a bit more open of a design, allowing you to buy the design of processor to integrate in your own chip. So many more manufacturers produce arm based processors, with various integrated peripherals.
Latest Answers