CUDA.jl
spack
CUDA.jl | spack | |
---|---|---|
15 | 52 | |
1,133 | 3,969 | |
1.1% | 1.6% | |
9.5 | 10.0 | |
7 days ago | 5 days ago | |
Julia | Python | |
GNU General Public License v3.0 or later | Apache-2.0 or MIT |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
CUDA.jl
-
Ask HN: Best way to learn GPU programming?
It would also mean learning Julia, but you can write GPU kernels in Julia and then compile for NVidia CUDA, AMD ROCm or IBM oneAPI.
https://juliagpu.org/
I've written CUDA kernels and I knew nothing about it going in.
- What's your main programming language?
-
How is Julia Performance with GPUs (for LLMs)?
See https://juliagpu.org/
-
Yann Lecun: ML would have advanced if other lang had been adopted versus Python
If you look at Julia open source projects you'll see that the projects tend to have a lot more contributors than the Python counterparts, even over smaller time periods. A package for defining statistical distributions has had 202 contributors (https://github.com/JuliaStats/Distributions.jl), etc. Julia Base even has had over 1,300 contributors (https://github.com/JuliaLang/julia) which is quite a lot for a core language, and that's mostly because the majority of the core is in Julia itself.
This is one of the things that was noted quite a bit at this SIAM CSE conference, that Julia development tends to have a lot more code reuse than other ecosystems like Python. For example, the various machine learning libraries like Flux.jl and Lux.jl share a lot of layer intrinsics in NNlib.jl (https://github.com/FluxML/NNlib.jl), the same GPU libraries (https://github.com/JuliaGPU/CUDA.jl), the same automatic differentiation library (https://github.com/FluxML/Zygote.jl), and of course the same JIT compiler (Julia itself). These two libraries are far enough apart that people say "Flux is to PyTorch as Lux is to JAX/flax", but while in the Python world those share almost 0 code or implementation, in the Julia world they share >90% of the core internals but have different higher levels APIs.
If one hasn't participated in this space it's a bit hard to fathom how much code reuse goes on and how that is influenced by the design of multiple dispatch. This is one of the reasons there is so much cohesion in the community since it doesn't matter if one person is an ecologist and the other is a financial engineer, you may both be contributing to the same library like Distances.jl just adding a distance function which is then used in thousands of places. With the Python ecosystem you tend to have a lot more "megapackages", PyTorch, SciPy, etc. where the barrier to entry is generally a lot higher (and sometimes requires handling the build systems, fun times). But in the Julia ecosystem you have a lot of core development happening in "small" but central libraries, like Distances.jl or Distributions.jl, which are simple enough for an undergrad to get productive in a week but is then used everywhere (Distributions.jl for example is used in every statistics package, and definitions of prior distributions for Turing.jl's probabilistic programming language, etc.).
-
C++ is making me depressed / CUDA question
If you just want to do some numerical code that requires linear algebra and GPU, your best bet would be Julia or Python+JAX.
-
Parallélisation distribuée presque triviale d’applications GPU et CPU basées sur des Stencils avec…
GitHub - JuliaGPU/CUDA.jl: CUDA programming in Julia.
- Why Fortran is easy to learn
-
Generic GPU Kernels
Should have (2017) in the title.
Indeed cool to program julia directly on the GPU and Julia on GPU and this has further evolved since then, see https://juliagpu.org/
-
Announcing The Rust CUDA Project; An ecosystem of crates and tools for writing and executing extremely fast GPU code fully in Rust
I'm excited to eventually see something like JuliaGPU with support for multiple backends.
-
[Media] 100% Rust path tracer running on CPU, GPU (CUDA), and OptiX (for denoising) using one of my upcoming projects. There is no C/C++ code at all, the program shares a single rust crate for the core raytracer and uses rust for the viewer and renderer.
That's really cool! Have you looked at CUDA.jl for the Julia language? Maybe you could take some ideas from there. I am pretty sure it does the same thing you do here, and they support any arbitrary code with the limitations that you cannot allocate memory, I/O is disallowed, and badly-typed code(dynamic) will not compile.
spack
-
Autodafe: "freeing your freeing your project from the clammy grip of autotools."
> Are we talking about the same autotools?
Yes. Instead of figuring out how to do something particular with every single software package, I can do a --with-foo or --without-bar or --prefix=/opt/baz-1.2.3, and be fairly confident that it will work the way I want.
Certainly with package managers or (FreeBSD) Ports a lot is taken care of behind the scenes, but the above would also help the package/port maintainers as well. Lately I've been using Spack for special-needs compiles, but maintainer ease also helps there, but there are still cases one a 'fully manual' compile is still done.
> Suffice it to say, I prefer to work with handwritten makefiles.
Having everyone 'roll their own' system would probably be worse, because any "mysteriously failure" then has to be debugged specially for each project.
Have you tried Spack?
* https://spack.io
* https://spack.readthedocs.io/en/latest/
-
FreeBSD has a(nother) new C compiler: Intel oneAPI DPC++/C++
Well, good luck with that, cause it's broken.
Previous release miscompiled Python [1]
Current release miscompiles bison [2]
[1] https://github.com/spack/spack/issues/38724
[2] https://github.com/spack/spack/issues/37172#issuecomment-181...
-
Essential Command Line Tools for Developers
gh is available via Homebrew, MacPorts, Conda, Spack, Webi, and as a…
-
The Curious Case of MD5
> I can't count the number of times I've seen people say "md5 is fine for use case xyz" where in some counterintuitive way it wasn't fine.
I can count many more times that people told me that md5 was "broken" for file verification when, in fact, it never has been.
My main gripe with the article is that it portrays the entire legal profession as "backwards" and "deeply negligent" when they're not actually doing anything unsafe -- or even likely to be unsafe. And "tech" knows better. Much of tech, it would seem, has no idea about the use cases and why one might be safe or not. They just know something's "broken" -- so, clearly, we should update.
> Just use a safe one, even if you think you "don't need it".
Here's me switching 5,700 or so hashes from md5 to sha256 in 2019: https://github.com/spack/spack/pull/13185
Did I need it? No. Am I "compliant"? Yes.
Really, though, the main tangible benefit was that it saved me having to respond to questions and uninformed criticism from people unnecessarily worried about md5 checksums.
- Spack Package Manager v0.21.0
- Show HN: FlakeHub – Discover and publish Nix flakes
-
Nixhub: Search Historical Versions of Nix Packages
[1] https://github.com/spack/spack/blob/develop/var/spack/repos/...
-
Cython 3.0 Released
In Spack [1] we can express all these constraints for the dependency solver, and we also try to always re-cythonize sources. The latter is because bundled cythonized files are sometimes forward incompatible with Python, so it's better to just regenerate those with an up to date cython.
[1] https://github.com/spack/spack/
-
Linux server for physics simulations
You want to look at the tools used for HPC systems, these are generally very well tried and tested and can be setup for single machine usage. Remote access - we use ssh, but web interfaces such as Open On Demand exist - https://openondemand.org/. For managing Jobs, Slurm is currently the most popular option - https://slurm.schedmd.com/documentation.html. For a module system (to load software and libraries per user), Spack is a great - https://spack.io/. You might also want to consider containerisation options, https://apptainer.org/ is a good option.
-
Simplest way to get latest gcc for any platform ?
git clone https://github.com/spack/spack.git ./spack/bin/spack install gcc
What are some alternatives?
LoopVectorization.jl - Macro(s) for vectorizing loops.
HomeBrew - 🍺 The missing package manager for macOS (or Linux)
cunumeric - An Aspiring Drop-In Replacement for NumPy at Scale
nixpkgs - Nix Packages collection & NixOS
awesome-quant - A curated list of insanely awesome libraries, packages and resources for Quants (Quantitative Finance)
nix-processmgmt - Experimental Nix-based process management framework
cudf - cuDF - GPU DataFrame Library
Ansible - Ansible is a radically simple IT automation platform that makes your applications and systems easier to deploy and maintain. Automate everything from code deployment to network configuration to cloud management, in a language that approaches plain English, using SSH, with no agents to install on remote systems. https://docs.ansible.com.
Tullio.jl - ⅀
ohpc - OpenHPC Integration, Packaging, and Test Repo
GPUCompiler.jl - Reusable compiler infrastructure for Julia GPU backends.
NixOS-docker - DEPRECATED! Dockerfiles to package Nix in a minimal docker container