Yesterday, Apple unveiled its latest marvel of computing technology, and no, I'm not talking about its new $129 pencil or its new $299 keyboard. It's not a new electric car that charges from the bottom either. It's the new M4 chip, which Apple claims is the biggest and baddest chip the PC market has ever seen. Its CPU is 50% faster than the M2, it's got built-in ray tracing, and the neural engine can handle a ridiculous 38 trillion operations per second, making it the fastest chip for AI workloads on the PC market today.

What's interesting, though, is that Apple chose to unveil the M4 yesterday at its iPad event and not its Worldwide Developer Conference, which is only a month away. What you may not realize is that we're in the middle of an arms race—a global race to build the best ARM-based chip and dominate the world. But Apple's lead may not last very long. In today's video, we'll look at the major players and what it all means for the future of computing technology. It is May 8th, 2024, and you're watching The Code Report.

What is ARM?

Everybody I know is building ARM chips. Samsung, Google, Microsoft, and even Intel are joining the arms race. But wait a second, what even is ARM? Historically, personal computers like those running Windows or macOS run on CPUs that use the x86 architecture, and those chips are typically built by Intel or AMD. x86 is based on the complex instruction set, which is extremely versatile and fast for a wide variety of computing needs. However, when smartphones became a thing, they chose to use ARM processors based on the reduced instruction set, which requires fewer transistors per instruction, making it more power-efficient. And that, of course, is critical for battery-powered devices.

But there was always an assumption that ARM wasn't optimal for regular computers. Well, that changed in 2020 when Apple released the M1 chip. Suddenly, their laptops were faster, quieter, and no longer caused third-degree burns on my crotal region. This motivated Microsoft to launch Project Volterra to run Windows on ARM chips. In addition, data centers like AWS have realized that ARM chips are more efficient in the cloud, and now you have Amazon, Google, and Microsoft all making their own ARM chips for their data centers, like AWS Graviton and Google Axion.

What's kind of crazy, though, is that the company behind this architecture is worth over $100 billion but has never actually made a single chip. Instead, it just designs the architecture, which is then licensed by big companies and ultimately manufactured at places like Taiwan Semiconductor. They produce Apple's M chips, and one big selling point of the M4 is that it has the most powerful neural engine ever. Apple's been very quiet in the AI race, but it looks like their goal is to run large language models on-device or on the edge. This would be awesome for users because then an LLM could work offline with low latency and, most importantly, keep all the chat messages with your AI girlfriend private on-device.

Why Apple?

Apple's lead may not last very long. In just a couple of weeks, Qualcomm is expected to release the Snapdragon X Elite chip, which could outperform the M4 and would be the chip running Windows on Surface laptops. They claim it can do 45 trillion operations per second versus a measly 38 trillion operations per second on the M4, although they might be cheating on these benchmarks. Now, the reason they're competing so fiercely on these numbers is that in the near future, the device that can run the best AI model natively is the device that people will buy. Apple is in the business of selling hardware, and strategically, it would be pretty embarrassing if Microsoft laptops could run AI better than Apple.

Is x86 Dead?

With all this talk about ARM, you might be wondering, is x86 now a dead technology? Well, some people out there make a pretty good argument that x86 needs to die. But keep in mind, it's been around for over 45 years, and it's so deeply embedded in modern infrastructure that it's likely going to outlive everyone here on planet Earth today. In addition, just the other day, documents were leaked about Intel's new Arrow Lake processor, which has been built from the ground up to consume less power and could potentially handle over 100 trillion operations per second. In addition, AMD is also coming out with the new Ryzen AI 9.

Now we're starting to see a pattern emerge here. The company that can maximize the AI cringe level on their branding is the company that will ultimately win this race. And that's why next month at WWDC, it's rumored that Apple will finally unveil the MacBook Singularity Blaster Turbo Pro Max. This has been The Code Report. Thanks for watching, and I will see you in the next one.