My notes on Apple’s M1 Chip

There are already many (and will be many more) recaps about Apple’s One More Thing keynote. The company unveiled its silicon, the M1 Chip, which will replace Intel processors in its lineup. For now, the M1 Chip will become the brains and heart of three consumer-focused Mac-devices: the Mac mini, the Macbook Air, and the Macbook Pro (the latter two both with 13-inch displays). Here are some notes I jotted down in my notebook about the keynote and the new M1 Chip:

First-time Mac buyers are buying one out of two new Macs sold.
The Mac business grew by nearly 30% last quarter, and the Mac is having its best year ever.

The M1 has 16 billion transistors — or as my friend, Steve says, two for every human on the planet. If I read this correctly, this is about a third larger than the latest and greatest A14 chip that powers the new iPhone12Pros.

The M1 has 4 performance cores and 4 efficiency cores as part of its CPU. The M1 8-core graphic processing unit (GPU) takes up a gigantic amount of space on the chip. The cache is pretty sizeable and will certainly boost the overall performance.
Apple thinks the 16 GB memory is the sweet spot for the three models they are offering. No information on higher memory capacities, but I wouldn’t be surprised if higher-end models come with more memory. Nothing imminent, though.
If you are a developer and want access to the same data with the CPU and GPU, you don’t have to copy it back and forth over PCI Express. It enables better performance for the application.

Games and applications that take advantage of Metal will perform better on the new platform, including ones that were written for Intel-based Macs and run using the Rosetta 2 translation layer. The integrated GPU will have more and better access to memory and give the requisite bandwidth boost.
The external eGPU units will not work with the M1 Chip-powered machines, and Apple believes the integrated graphic processor capabilities are enough to outperform the discrete GPUs.
Apps that are reliant heavily on GPU will probably be better off sticking to higher-end Macs. Apple insists that the M1 can do a better job and has the best-integrated GPU.
The secure enclave is on M1, so there is no T2 security chip in Macs.
It is not clear which Intel chips are being used compared to M1, but if you take a cue from Qualcomm’s similar chips, the iCore5 Chip is a likely comparison.
Apple isn’t talking about clock speeds.
Apple will push on performance per watt and tout better battery life as a critical point of differentiation.
M1 is the start of a long transition and will bring significant advantage to the company, especially when compared to its x86 based rivals.

The Bottom Line: Given how tightly integrated the software and hardware are at Apple, the new silicon is going to surprise us all. That said, Apple has a significant marketing challenge. While the iPhone was brand new and allowed Apple to write the narrative around performance. Apple’s biggest challenge is around marketing — we humans are conditioned to thinking about laptops and desktop computers in terms of clock speeds, memory, and performance.