M3 chips widen the gap between Pro and Max variants. They also change relative performance between P and E cores to make M3 CPUs more versatile.
performance
The M1 cycle took 16 months from basic to Ultra; that shortened to 12 months for the M2. As the first Studio M2 Ultras were being prepared for shipping, the M3 cycle started.
A strange observation, that the last thread to complete a matrix multiplication task was always much later than others, explored to discover a different strategy used by macOS.
Comparison between 2 Intel and 2 Apple silicon Macs running vector and matrix functions from Apple’s Accelerate library. Was that new M3 worth the money?
There’s more to getting best performance and energy efficiency on Apple silicon. These vary greatly depending on how apps are coded, as shown here.
If Apple offered to do much of the hard work of coding your app for you for free, and to optimise it for different Mac hardware, how could you refuse?
How to compare an undocumented if not secret co-processor? Using different tests that use very high power, and can result in strange patterns of core allocation. So how does the M3 Pro fare here?
Comparison with M1 variants, energy use with comparison between M3 Pro and Max, virtualisation, Game Mode, vector processing and matrix co-processing – all in summary.
Assessing throughput using tests of fast Fourier transforms and sparse Cholesky factorisation from the Accelerate library. Is there an AMX there?
Differences in vector processing performance between the M1 Max and M3 Pro, and in their use of power. Their frequency control is more complex.
The Eclectic Light Company 