#pythran Suchergebnisse
Milestone reached: we've got 256 stars for #xsimd. That means we've got as many stars on xsimd as there are bits in a AVX register! In other news: we've got all integer types for AVX and SSE and, the big one, #pythran switched to xsimd from boost.SIMD and all tests are passing!
Check out the @ProjectJupyter magic for compiling #python code with #pythran in a notebook Cell, and the improved performance over #numpy!
Activating boost.SIMD in #pythran, and updating to the master branch yields incredible speedups! #xtensor will have to keep catching up for a while longer 😏
Serge Guelton presenting of the #pythran Python to C++ transpiler at the @OpenDreamKit low-level interfaces workshop.
À travers le projet Pythran, Serge Guelton nous montre des résultats prometteurs 👍 #Pythran #PyData #Python
I've forked @SimonDanisch 's nextjournal article to include a comparison of the speed of R, @graalvm, @JuliaLanguage and #Pythran (Python-to-C++ compiler). Julia and Pythran are equally fast! nextjournal.com/wolfv/how-fast… Shoutout to @usenextjournal it's nice & easy to mix languages!
First day of the open-source scientific computing workshop on low-level interfaces with core developers of @sagemath @ProjectJupyter #cython #pythran #xtensor @OpenDreamKit.
Forked the awesome #pythran @numba_jit #numpy benchmarks from Serge (the pythran author) to add some xtensor-python benchmarks! Results look good! github.com/wolfv/numpy-be… #xtensor
The PR in #pythran that removed 160k and added only 30k lines of code: github.com/serge-sans-pai…
🚀 Python too slow for science? Try Pythran! It magically compiles Python→C++ for blazing speed—no code changes! Just add #pythran hints. open.substack.com/pub/pythonlibr…
@pydataparis, code faster at a higher level, réduce engineering cost #pythran to translate python efficiently #c++ integration
1. #PyPy, #Numba, #Pythran & #Transonic with some changes in the original code improves performance remarkably. 2. Reduction in time to solution compensates for increased power requirement for multithreading. Article: rdcu.be/ciO0J Code:… pythondevs.social/notice/A6ZhFgG…
Pythran just released its long-awaited update: 0.9.7. Here's the changelog: github.com/serge-sans-pai… #pythran
So I have done this and I do not regret it. I am pretty optimistic now the two projects will overlap. I need to get #nuitka further ahead with C type inference over module boundary, but with hacks, it will work with #pythran, looking forward to meeting!
The (pure Python) source code is here: gist.github.com/wolfv/cecbfc35… Note that to get these speeds, only the small #Pythran comment is necessary that tells the pythran compiler the shape and type of the numpy input.
xsimd 7.5.0 is out! Tons of improvements and new features in this release: - saturated arithmetics - basic shuffle operations - support for AVX512 on Windows The author of #Pythran is now a core contributor to the project, recently adopted by #ApacheArrow. github.com/xtensor-stack/…
github.com
GitHub - xtensor-stack/xsimd: C++ wrappers for SIMD intrinsics and parallelized, optimized mathem...
C++ wrappers for SIMD intrinsics and parallelized, optimized mathematical functions (SSE, AVX, AVX512, NEON, SVE)) - xtensor-stack/xsimd
#pythran 2.7 (very fast Python2.7->C++->compiled module) released, you 'do no work and it runs faster': pythonhosted.org/pythran/ #python
🚀 Python too slow for science? Try Pythran! It magically compiles Python→C++ for blazing speed—no code changes! Just add #pythran hints. open.substack.com/pub/pythonlibr…
Thanks #Pythran people for being so polite: WARNING: Compilation error, trying hard to find its origin... WARNING: Nop, I’m going to flood you with C++ errors!
#Pythran is an ahead of time compiler for a subset of the #Python language, with a focus on #ScientificComputing pythran.readthedocs.io/en/latest/ #AI #DataScience #Analytics #DataAnalyisis #libatlas #numpy
Had a great time at #SciPy2021 yesterday, best virtual conference experience I've had so far! Gave a talk together with Serge Guelton on using #Pythran in #SciPy to make it easier to write fast kernels, here are the slides if you're curious: slideshare.net/RalfGommers/bu….
slideshare.net
Building SciPy kernels with Pythran
1. #PyPy, #Numba, #Pythran & #Transonic with some changes in the original code improves performance remarkably. 2. Reduction in time to solution compensates for increased power requirement for multithreading. Article: rdcu.be/ciO0J Code:… pythondevs.social/notice/A6ZhFgG…
"Reducing the ecological impact of computing through education and Python compilers". Excellent read by @pierre_augier, @cfbolz, and Serge Guelton, in @nature. Featuring the #Pythran open-source project! nature.com/articles/s4155…
nature.com
Reducing the ecological impact of computing through education and Python compilers
Nature Astronomy - Reducing the ecological impact of computing through education and Python compilers
Do not hesitate to have a look at github.com/xtensor-stack/…. Not saying you should switch to it,but you could pick up ideas there. It is well maintained and used by #Xtensor #Pythran and #ApacheArrow.
github.com
GitHub - xtensor-stack/xsimd: C++ wrappers for SIMD intrinsics and parallelized, optimized mathem...
C++ wrappers for SIMD intrinsics and parallelized, optimized mathematical functions (SSE, AVX, AVX512, NEON, SVE)) - xtensor-stack/xsimd
Oh, I was just working on a benchmark experiment on the n-body planets algorithm, in particular with Python implementations (using different accelerators/interpreters, #Transonic #Pythran @numba_jit @pypyproject), see github.com/paugier/nbabel
github.com
GitHub - paugier/nbabel
Contribute to paugier/nbabel development by creating an account on GitHub.
Pythran just released its long-awaited update: 0.9.7. Here's the changelog: github.com/serge-sans-pai… #pythran
Basic Value Range Analysis #pythran serge-sans-paille.github.io/pythran-storie…
Did you know you easily install the @vaex_io DataFrame library on @kaggle kernels and accelerate computations via @numba_jit , #Pythran and #CUDA. On their GPU instances @kaggle use @nvidia Tesla P100 GPUs, with a dual core CPU. (1/3)
Are the slides available anywhere? I'd love to read them! Were #Pythran and #Transonic mentioned? People interested by the subject (Tools for high performance Python) should read this text tiny.cc/transonic-visi… @ianozsvald #PyDataCambridge
@numba_jit #cython #pythran @scikit_learn @scikitbio @statsmodels @pymc_devs @astropy @Biopython @SunPyProject @poliastro_py #scikitimage @jnuneziglesias @pyOpenSci @SciPyTip
Transonic 0.4.0 is on PyPI and @condaforge. Much better than the previous version legi.grenoble-inp.fr/people/Pierre.… ! Easily accelerate #Python #NumPy code with @numba_jit #Pythran and #Cython. transonic.readthedocs.io @pyfluiddyn
Set of tutorials for high performance computing with python, applied to crystallography @SciPyTip @esrfsynchrotron #python #numexpr #pythran #cython #numba journals.iucr.org/j/issues/2019/…
Great slides! For HPC with Python/Numpy, don't forget #pythran, which can be much faster than #numba. Super easy to use with transonic.readthedocs.io Just few `@boost` and `@jit` can make numerical codes super efficient.
We intend to integrate with #pythran at some point. About Numba I have no clue, but it might be a lot better at this point if you do not need full Python compatibility, but seriously I don't know. Mid term I would expect us to overtake what they do once we do PGO and JIT too.
Beniget is used by Transonic (transonic.readthedocs.io), a pure Python package to accelerate Python-Numpy numerical kernels with different accelerators, in particular #pythran. Soon compatible with @numba_jit and #cython !
Check out the @ProjectJupyter magic for compiling #python code with #pythran in a notebook Cell, and the improved performance over #numpy!
Activating boost.SIMD in #pythran, and updating to the master branch yields incredible speedups! #xtensor will have to keep catching up for a while longer 😏
Serge Guelton presenting of the #pythran Python to C++ transpiler at the @OpenDreamKit low-level interfaces workshop.
Milestone reached: we've got 256 stars for #xsimd. That means we've got as many stars on xsimd as there are bits in a AVX register! In other news: we've got all integer types for AVX and SSE and, the big one, #pythran switched to xsimd from boost.SIMD and all tests are passing!
First day of the open-source scientific computing workshop on low-level interfaces with core developers of @sagemath @ProjectJupyter #cython #pythran #xtensor @OpenDreamKit.
À travers le projet Pythran, Serge Guelton nous montre des résultats prometteurs 👍 #Pythran #PyData #Python
Forked the awesome #pythran @numba_jit #numpy benchmarks from Serge (the pythran author) to add some xtensor-python benchmarks! Results look good! github.com/wolfv/numpy-be… #xtensor
I've forked @SimonDanisch 's nextjournal article to include a comparison of the speed of R, @graalvm, @JuliaLanguage and #Pythran (Python-to-C++ compiler). Julia and Pythran are equally fast! nextjournal.com/wolfv/how-fast… Shoutout to @usenextjournal it's nice & easy to mix languages!
1. #PyPy, #Numba, #Pythran & #Transonic with some changes in the original code improves performance remarkably. 2. Reduction in time to solution compensates for increased power requirement for multithreading. Article: rdcu.be/ciO0J Code:… pythondevs.social/notice/A6ZhFgG…
Something went wrong.
Something went wrong.
United States Trends
- 1. Columbus 65.8K posts
- 2. #SwiftDay 8,325 posts
- 3. #WWERaw 20.7K posts
- 4. #IDontWantToOverreactBUT N/A
- 5. #IndigenousPeoplesDay 2,567 posts
- 6. #TSTheErasTour 2,111 posts
- 7. $ZOOZ 1,021 posts
- 8. Knesset 115K posts
- 9. Marc 36.3K posts
- 10. Good Monday 39.3K posts
- 11. Victory Monday 1,452 posts
- 12. Thanksgiving 43.5K posts
- 13. Flip 49.7K posts
- 14. Kairi 9,745 posts
- 15. Branch 48.1K posts
- 16. Broadcom 1,608 posts
- 17. Penta 4,753 posts
- 18. Rod Wave 3,057 posts
- 19. Happy 250th 1,886 posts
- 20. The Final Show 26.4K posts