Early tests show a on pure-Python numeric loops and repeated function calls—without any code changes. The JIT is disabled by default; enable it via a special build flag.
October 7, 2024 – The Python community today celebrates the stable release of Python 3.13 , a version that its developers are calling one of the most intriguing updates in recent memory. While not a full-speed revolution, 3.13 plants the seeds for a dramatically faster future—and gives developers powerful new tools to play with today. python 3.13 news today
Here’s what’s new, what’s experimental, and what disappears. For decades, Python’s Global Interpreter Lock (GIL) has been a controversial cornerstone. It simplifies memory management but prevents true parallel execution of threads. In Python 3.13, the GIL is still the default —but for the first time, you can compile a Python interpreter without it. Early tests show a on pure-Python numeric loops
This mode (enabled via --disable-gil at build time) allows multiple threads to run Python code simultaneously on multiple CPU cores. The result? True parallelism for CPU-bound tasks without resorting to multiprocessing. While not a full-speed revolution, 3
It’s experimental. Some C extensions may break, and single-threaded performance takes a small hit (roughly 10% slower). However, for scientific computing, web servers, and data processing, early benchmarks show impressive gains on multi-core machines. “This is not for production just yet,” said a core developer in the release notes, “but we need users to try it, break it, and report back. This is how we prepare for Python 3.14 or 3.15.” JIT Compilation: The Quiet Game-Changer Python 3.13 quietly introduces an experimental Just-In-Time (JIT) compiler . Unlike full JITs in languages like Java or JavaScript, Python’s initial implementation is modest: it compiles bytecode to machine code for small, hot regions of code.
As one release note put it: “Python 3.13 doesn’t change how you write code. It changes how your code runs. Try it, break it, and help us build a faster Python for tomorrow.” End of story