In late 2020, Apple unveiled its first M1 system-on-a-chip (SoC), which combines Apple’s custom ARM-based CPUs with its GPUs, Neural Engine for AI applications, Secure Enclave, and various signal and signal processors. A driver for the I/O interface. In addition to the MacBook Air and 13-inch MacBook Pro, the chip is also found in the Mac mini. The higher-performance M1 Pro, Max, and Ultra models have since been released, and the SoC can now be found in Apple’s laptops and desktops, including the updated Mac Studio for professional users.

Perhaps it’s no surprise that Apple is now using its WWDC conference to showcase its next generation of Apple Silicon. This is a year and a half after first announcing the M1. M2 was expected to be shown off at Apple’s latest hardware event, but instead, we got the M1 Max and Ultra.

Apple noted that when designing the M2, it focused on power efficiency. The chip uses 5nm technology and has an 8-core CPU and a 10-core GPU for a total of 20 billion transistors. The memory controller supports up to 24 GB of memory and has a unified memory bandwidth of 100 GB/s, 50% more than M1.

Apple also compared the performance and power consumption of the M2 chip to a 12-core PC chip. According to the company, the M2 uses a quarter of the power, resulting in the same performance. Even though the CPU is “only” 18% faster than the M1, it’s still a huge improvement in the CPU business, where performance gains these days tend to be in the single digits. As for the GPU, the M2 now has 10 cores, two more than the M1. At the same power level as the M1, this gives you 25% more performance and 35% more at maximum performance.