Python 3.15 JIT Compiler Achieves Performance Goals Ahead of Schedule
By
guidoiaquinti
An everything bagel for the brain. Substantive, layered, well-seasoned.
Summary
The article reports that Python 3.15's JIT compiler development is back on track and has achieved its performance goals ahead of schedule. The JIT shows 11-12% speed improvements on macOS AArch64 and 5-6% improvements on x86_64 Linux compared to interpreters, though performance varies across different benchmarks. The article discusses the technical challenges, implementation details, and future roadmap for the CPython JIT project.
Key quotes
· 5 pulledGreat news—we've hit our (very modest) performance goals for the CPython JIT over a year early for macOS AArch64, and a few months early for x86_64 Linux.
The 3.15 alpha JIT is about 11-12% faster on macOS AArch64 than the tail calling interpreter, and 5-6% faster than the standard interpreter on x86_64 Linux.
These numbers are geometric means and are preliminary. The actual range is something like a 20% slowdown to a 30% speedup depending on the benchmark.
The JIT is now back on track after some delays and technical challenges.
We're focusing on stability and correctness first, then we'll work on further performance optimizations.
You might also wanna read
Python's splitlines() Method Handles More Line Boundaries Than Just Newlines
The article reveals that Python's str.splitlines() method splits strings on more line boundaries than just the common newline characters (\n
Rethinking Design Patterns in Python: Why Some Are Unnecessary
The article critiques the overuse of traditional design patterns in Python, arguing that many of these patterns, originally designed for lan
lihil.cc·10mo ago
Why Average LLM Use Is Likely Destroying Value in Software Development
The author argues that, contrary to prevailing hype, the average use of Large Language Models (LLMs) is likely destroying value rather than
unessays.substack.com·10h ago
How AI Accelerated Prototyping: From Idea to Tangible in Record Time
The author reflects on how AI has transformed their prototyping workflow. Previously, the biggest bottleneck was the time needed to scaffold
darylcecile.net·10h ago
GitLab 19.0 launches with Secrets Manager, agentic workflows, and self-hosted AI models
GitLab 19.0 has been released, positioning itself as an intelligent orchestration platform for DevSecOps. The release includes expanded secr
bit.ly·23h ago
Centralizing Error Handling in Rust with Custom AppError Enums
This article discusses the importance of centralizing error handling in Rust applications using a custom AppError enum combined with map_err