HN · HACKER_NEWS// 25d agoPRODUCT UPDATE

Python 3.15 JIT hits double-digit speed gains

The CPython Just-In-Time (JIT) compiler has reached significant performance milestones for the upcoming Python 3.15 release, achieving an 11-12% speedup on macOS AArch64. Following a transition to community stewardship in 2025, the project successfully overcame earlier performance hurdles through technical innovations like dual-dispatch trace recording and optimized reference count elimination.

// ANALYSIS

Python 3.15’s JIT is a watershed moment for CPython, proving that a lean, community-led effort can deliver where corporate-sponsored initiatives hit walls.

–The new "dual dispatch" tracing mechanism significantly improves JIT code coverage by 50% without penalizing the base interpreter's performance.
–Speed gains are highly platform-dependent, with Apple Silicon users seeing double-digit improvements while x86-64 Linux gains hover around 5-6%.
–While not a direct replacement for C-extensions or NumPy, it provides a "zero-effort" performance floor raise for general-purpose logic and orchestration code.
–The project's roadmap now targets free-threading compatibility, which will be the ultimate test for the JIT's impact on high-concurrency AI workloads.

// TAGS

pythonjitopen-sourcedevtoolbenchmark

DISCOVERED

25d ago

2026-03-17

PUBLISHED

25d ago

2026-03-17

RELEVANCE

7/ 10

AUTHOR

guidoiaquinti