Tag
Apple announces the M5 chip designed for server applications, expanding its custom silicon line for data centers.
Apple M5 silicon supports INT8 activations for matrix multiplication, but inference backends like MLX and Llama.cpp currently use 16-bit; custom w8a8 kernels achieve up to 1.4x speedup on Gemma4 prefill tasks.
A comprehensive buying guide comparing Apple's latest MacBook models (Neo, Air, Pro) and their M5, M4, and older chips, helping readers choose the right device.