I wrote a previous article on my experience to date with Apple’s new M1 based MacBook Pro. There I discussed why I thought it was so great and a few concerns. Overall verdict: Awesome!
But first – a quick plug for the Erie Island Coffee Company – where I have found a Covid-19 home to work that is fun and productive, and where this picture was taken.
I provide this article for those who want to know why it is a game-changer and what Apple did differently. It goes without saying, but I am not in the league of the smart people that developed or documented Apple’s approach, but there is enough info on the web to understand why it matters from my perspective, and perhaps yours.
WHY is the Apple MacBook Pro M1 So Much Faster?
Some history will be helpful to understand the transition. Apple is one of the oldest personal computer makers and has gone through several generations of technologies. Most were the typical “this year’s chip, same as last year’s, but faster.” But on four occasions, they have wholly switched platforms to a different chip architecture.
- Apple started in 1976 with the MOS technology 6502 processor, an eight-bit chip found mostly in the Apple II line.
- In 1984 they unveiled the Apple Macintosh line, based on a Motorola 68000 CPU, a 16/32 bit chip.
- In 1994, Apple switched the Macintosh to a PowerPC CPU, developed jointly by Apple, Motorola, and IBM.
- In 2005, Apple moved the Macintosh line to Intel processors, often called X86 and known by names like 80486, Pentium, and I7. The PowerPC CPU had fallen behind compared to chips from Intel and AMD, and Apple was suffering as a result. The shift also allowed Mac PCs to run Microsoft Windows directly or in a virtual machine.
- At the end of 2020, Apple released new personal computer devices (MacBook Air, MacBook Pro, Mac Mini) based on their custom chipset called M1. This chip has many advantages over the Intel-based chips (still available from Apple but will be phased out eventually).
Why all the History of the Chips?
Because the change is not just evolutionary or faster of the same, it is revolutionary in a sense. It is a different approach. The M1 CPU is a conglomeration of functions, tightly bound together and with a better instruction set; this constitutes a “System On a Chip” or “SOC.” Traditional computers have a motherboard with many chips like the CPU, memory, input/output, and many more; the SOC combines these. The M1 SOC is better described here by someone smarter than me, so read it if you want to understand in detail but let me summarize herein.
There are multiple functions the SOC does:
- CPU – The actual place where it executes instructions, like adding numbers together. The CPU on the M1 is the same type as in a current iPad or iPhone (the A14 Bionic). It has eight cores, meaning it can process eight requests simultaneously.
- Graphics – the graphics requirements in a PC are usually carried out by a separate module, like a graphics card (sometimes called a GPU). These cards can be quite expensive, noisy, and power-hungry. The Apple M1 boasts eight graphical cores tightly integrated to the CPU, so it quickly transfers instructions to the GPU, and the whole process takes less time, generates less heat, and requires less power.
- Memory – like the traditional GPU, memory is usually somewhere else in the computer connected by lines on a motherboard. Additionally, the CPU has memory on the motherboard, and the GPU has its memory. And when the CPU wants to share something from its memory with the GPU, it must send it, which takes time (superfast, but takes time and happens continuously). On the M1, the memory is unified, so when the CPU has memory it wants to share with the GPU, the GPU already has it. Now you see how it all starts to work together!
- More functions are built-in: Neural processing (for Machine Language AI), image processing, digital signal processing (for things like decoding a music file), and secure enclave (for encryption and other things). And once again, these do not require a trip across the motherboard.
Apple took its superfast A14 Bionic chip in the iPhone and iPad, combined it with some other vital needs, and put it all in one place. That makes a big difference.
What else is different?
The other key difference from the X86 design is the instruction set. The M1 uses a version of the ARM processer, a Reduced Instruction Set Computing (RISC) design; the Intel model uses a Complex Instruction Set Computing (CISC) model. The difference is essential when considering the number of cores and SOC design, and something called decoders. The Intel CISC approach has more instructions it can process of varying length (from 1-15), allowing it to do more in one cycle of actions. The RISC approach of the Apple M1 processes a constant set of four instructions. Because these CPUs have multiple cores, it allows them to process multiple sets of instructions simultaneously. But instructions do not always come in the correct order, especially with a multi-tasking operating system running many applications simultaneously. CPUs handle this by using a decoder, which examines the sets of instructions and picks the next one to process. On Intel CPUs, they have up to four decoders examining instructions or various lengths; on the M1, eight decoders are processing instructions that are consistently four steps. The M1 design is simply more straightforward and faster.
The best example I can provide a box packing line. Which of these would be faster?
- Having four workers pack boxes with various size objects?
- Having eight workers pack boxes with the same size objects?
Option two is faster. And that is the real difference between the M1 and the Intel X86 approaches.
Why doesn’t the Wintel Gang emulate the Apple M1 MacBook Pro Design?
You may ask, when will Intel or AMD and Microsoft (collectively: “Wintel”) do this same thing with Windows? Well, that is not as easy. Windows runs on a myriad of different chipsets and from numerous manufacturers. So, they try to develop in a way a manufacturer can adopt it and tweak it for some market differentiator. But this causes Windows to be slower to change since everybody has to change, and they cannot tightly integrate the operating system with the chipset.
But Apple can because they sell you the hardware and operating system.
Secondly, the CISC to RISC move would be a significant one and cause problems for developers and IT departments who love stability.
Summary of Why the Apple MacBook Pro M1 is so Good!
It comes down to multiple reasons, all working in concert: the combination of various functions in one place, a consistent instruction set size, and an operating system tightly integrated with the hardware.
I find the MacBook Pro with the M1 chipset more productive than other devices for “real” work. The extra speed, better battery life, and overall ergonomics are excellent!
You can purchase the same model I have at Amazon – on sale now! Click here for affiliate link, and help support this site.
As always, contact us for help with enterprise or executive-level technology use. And follow on Twitter stay up to date.
Trackbacks/Pingbacks