Skip to content

Commit 52e3566

Browse files
Address Adam's review
1 parent ee16268 commit 52e3566

File tree

1 file changed

+10
-11
lines changed

1 file changed

+10
-11
lines changed

Doc/whatsnew/3.15.rst

Lines changed: 10 additions & 11 deletions
Original file line numberDiff line numberDiff line change
@@ -75,7 +75,8 @@ Summary -- Release highlights
7575
<whatsnew315-pep782>`
7676
* :ref:`Improved error messages <whatsnew315-improved-error-messages>`
7777
* The :ref:`JIT compiler <whatsnew315-jit>` has been significantly upgraded,
78-
with 3-4% geometric mean performance improvement on the pyperformance benchmarks.
78+
with 3-4% geometric mean performance improvement on the `pyperformance
79+
<https://github.com/python/pyperformance>`__ benchmarks.
7980

8081

8182
New features
@@ -857,16 +858,15 @@ csv
857858
Upgraded JIT Compiler
858859
=====================
859860

860-
The JIT compiler introduced in
861-
`Python 3.13 <https://docs.python.org/3.15/whatsnew/3.13.html#an-experimental-just-in-time-jit-compiler>`_
862-
has been significantly upgraded.
863-
864-
Current results from the pyperformance benchmark suite report
861+
Results from the pyperformance benchmark suite report
865862
`3-4% <https://github.com/facebookexperimental/free-threading-benchmarking/blob/main/results/bm-20251214-3.15.0a2%2B-6cddf04-JIT/bm-20251214-vultr-x86_64-python-6cddf04344a1e8ca9df5-3.15.0a2%2B-6cddf04-vs-base.svg>`_
866-
geometric mean performance over the standard CPython interpreter built with
867-
all optimizations enabled. The speedups range from roughly 20% slowdown to
868-
over 100% speedup (ignoring the ``unpack_sequence`` microbenchmark) on the
869-
x86-64 Linux and AArch64 macOS systems. These results are not yet final.
863+
geometric mean performance for the JIT over the standard CPython interpreter built with
864+
all optimizations enabled. The speedups for JIT builds versus no JIT builds
865+
range from roughly 20% slowdown to over 100% speedup (ignoring the
866+
``unpack_sequence`` microbenchmark) on x86-64 Linux and AArch64 macOS systems.
867+
868+
.. attention::
869+
These results are not yet final.
870870

871871
The major upgrades to the JIT are as follows:
872872

@@ -875,7 +875,6 @@ The major upgrades to the JIT are as follows:
875875
* Basic register allocation in the JIT
876876
* More JIT optimizations
877877
* Better machine code generation
878-
* ...watch this space for more to come before 3.15 beta!
879878

880879
LLVM 21 Build-Time Dependency
881880
-----------------------------

0 commit comments

Comments
 (0)