nogil multi-threading is slower than multi-threading with gil for CPU bound

[Venkat2811]

Bug report

Bug description:

Hello Team,

Thanks for the great work so far and recent nogil efforts. I wanted to explore performance of asyncio with CPU-GPU bound multi-threading in nogil setup. So, started with simple benchmark as seen below:

import sys
from concurrent.futures import ThreadPoolExecutor
def fib(n):
    if n < 2: return 1
    return fib(n-1) + fib(n-2)

threads = 8
if len(sys.argv) > 1:
    threads = int(sys.argv[1])

with ThreadPoolExecutor(max_workers=threads) as executor:
    for _ in range(threads):
        executor.submit(lambda: print(fib(34)))

Original Source

Results:

• python 3.13.a06 with gil:
  htop shows 2 running tasks, and only one core is utilized for 100%

$ time /usr/bin/python3.13 fib.py      
9227465
9227465
9227465
9227465
9227465
9227465
9227465
9227465
/usr/bin/python3.13 fib.py  6,46s user 0,04s system 100% cpu 6,447 total

• libpython3.13-nogil amd64 3.13.0~a6-1+jammy2 source nogil:
  htop shows 9 running tasks, and 8 cores close to 100% utilization, but slower.

$ time /usr/bin/python3.13-nogil -X gil=0 fib.py
9227465
9227465
9227465
9227465
9227465
9227465
9227465
9227465
/usr/bin/python3.13-nogil -X gil=0 fib.py  168,81s user 0,06s system 788% cpu 21,410 total

My CPU: AMD Ryzen 7 5800X 8-Core Processor
OS: Ubuntu 22.04.4 LTS

So looks like there is overhead when using multiple cores. Is this expected with this version ? Are results similar with Intel & M1 CPUs as well ?

Results documented here in version 3.9.12 on Intel is better.

CPython versions tested on:

3.13

Operating systems tested on:

Linux

[corona10]

FYI, the performance issue is not a surprise because we turned off the specialization, and deferred reference counting is not yet implemented for free-threading version.

But worth to checkout if there is unexpected bottleneck here.

[colesbury]

Please use 3.13 beta 1. There were a number of scaling bottlenecks fixed between 3.13 alpha 6 and beta 1 (#118527).

On my machine, I see a speedup in the free-threaded build vs. the default bulid (2.4s vs. 8.2s)

[corona10]

With ed2b0fb
number of thread: os.cpu_count()
task: fib(5)

[Venkat2811]

Thanks for confirming that things have been fixed in beta1 @colesbury

I'm waiting for beta1 build to be available on deadsnakes. I tried building from source, but unfortunately after trying several --with-ssl solutions, still getting not found errors.

[Venkat2811]

@colesbury I installed 3.13.0b1 via pyenv.

env PYTHON_CONFIGURE_OPTS='--disable-gil' pyenv install 3.13.0b1

Is this the way to create free-threaded build ?

On my M1 MBP:
--disable-gil vs default build: 15.5s vs 5.26s

On my AMD Ryzen 7 5800X 8-Core Processor, Ubuntu:
--disable-gil vs default build: 10.5s vs 6.4s

It has improved since 3.13.a06 for sure

[stonebig]

cpython-3.13.0b1 on Windows 11, for 20 threads, on a 4 cores/8 threads intel cpu: (faster is better)

• free-threading = 12 s
• standard = 19 s

[FeldrinH]

This is a slightly different benchmark, but there are some potentially useful data points about the speedup with free-threading here: winpython/winpython#1339.