But! Before you buy one of the still M4-based MacBook Airs, our tests of the MacBook Pro’s new M5 chip should give you an idea whether it’s worth waiting a few months (?) for an Air refresh.
Apple M5 review
We also ran a few M5 benchmarks as part of our M5 iPad Pro review, but running macOS rather than iPadOS gives us a lot more flexibility when it comes to testing: more benchmarks and a handful of high-end games to run, as well as command line access to examine power consumption and efficiency.
To back up and rephrase the chip specs for a moment, the M5 is built from the same building blocks as the M4: four high-performance CPU cores, six high-efficiency CPU cores (up from four in the M1/M2/M3), 10 GPU cores, and a 16-core Neural Engine to handle some machine learning and AI workloads.
The M5’s technical improvements are more focused and subtle than a simple increase in clock speed or core count. The first is a 27.5% increase in memory bandwidth, from the M4’s 120 GB/s to 153 GB/s (achieved, I’m told, by a combination of faster RAM and a memory structure that makes communication between different areas of the chip easier. Integrated GPUs are typically bottlenecked by memory bandwidth first and core count second, so improvements in bandwidth memory bandwidth can have a fairly direct and linear impact on graphics performance.
Apple also says it has added a “neural accelerator” to each of its GPU cores, separate from the neural engine. These will benefit a few specific types of workloads: Things like MetalFX graphics scaling or frame generation that would previously have had to use the Neural Engine can now do that work entirely within the GPU, eliminating some latency and freeing the Neural Engine to do other things. Apple also claims “4x more GPU compute than the M4,” which Apple says will accelerate AI language models and image generation software running locally. This figure mainly comes from GPU improvements; according to Geekbench AI, the Neural Engine itself is only about 10% faster than that of the M4.
