Apple today announced the M5 SoC as the newest in the Apple Silicon family and with claims of 4x the peak GPU compute performance for AI of the M4.
The Apple M5 brings more powerful CPU cores, a faster Neural Engine, and higher unified memory bandwidth, but their big focus was on the GPU performance for AI.
The Apple M5 is built on a third-generation 3nm process and features a next-gen 10-core GPU architecture with claims of being 4x the peak GPU compute performance of the M4.
On the CPU side, the Apple M5 features up to ten CPU cores between six efficiency cores and four performance cores. The multi-threaded performance is said to be a modest 15% improvement over the M4. The unified memory bandwidth for the M5 is up by nearly 30% to 153GB/s.
Initially the new Apple M5 is powering new models of the MacBook Pro, iPad Pro, and Vision Pro. More details on the Apple M5 at Apple.com.
Meanwhile the upstream Linux kernel support for Apple Silicon remains centered around the aging Apple M1/M2 SoCs and devices with the Apple M3/M4 bring-up still underway by the remaining Asahi Linux developers. The current Asahi Linux feature matrix for the prior-gen Apple Silicon hardware can be found via AsahiLinux.org.
