Mark Shannon’s “Shannon Plan” for Faster CPython outlines three parallel initiatives—Tier 2 optimizer, enabling sub‑interpreters via PEP 554, and memory‑management improvements—aimed at making Python up to five times faster by Python 3.13, with Microsoft joining the effort.
This article traces the history of the Python Global Interpreter Lock, explains why it remains a performance bottleneck, and examines recent efforts such as multiprocessing, PEP 554, and sub‑interpreters that aim to mitigate or eliminate the GIL in CPython.
Mark Shannon’s “Shannon Plan” and the Faster CPython project aim to boost Python’s speed fivefold within four years by introducing a Tier 2 optimizer, enabling sub‑interpreters via PEP 554, and overhauling memory management, while also discussing JIT prospects and internal debates over GIL removal.
The article outlines the Shannon Plan, a multi‑year effort led by CPython core developers to accelerate Python by introducing a Tier 2 optimizer, enabling sub‑interpreters via PEP 554, and improving memory management, with the ultimate goal of making Python five times faster by version 3.13.
