LoopVectorization.jl
CUDA.jl
LoopVectorization.jl | CUDA.jl | |
---|---|---|
10 | 15 | |
722 | 1,137 | |
0.6% | 1.5% | |
7.0 | 9.5 | |
5 days ago | 1 day ago | |
Julia | Julia | |
MIT License | GNU General Public License v3.0 or later |
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.
LoopVectorization.jl
-
Mojo – a new programming language for all AI developers
It is a little disappointing that they're setting the bar against vanilla Python in their comparisons. While I'm sure they have put massive engineering effort into their ML compiler, the demos they showed of matmul are not that impressive in an absolute sense; with the analogous Julia code, making use of [LoopVectorization.jl](https://github.com/JuliaSIMD/LoopVectorization.jl) to automatically choose good defaults for vectorization, etc...
```
- Knight’s Landing: Atom with AVX-512
-
Python 3.11 is 25% faster than 3.10 on average
> My mistake in retrospect was using small arrays as part of a struct, which being immutable got replaced at each time step with a new struct requiring new arrays to be allocated and initialized. I would not have done that in c++, but julia puts my brain in matlab mode.
I see. Yes, it's an interesting design space where Julia makes both heap and stack allocations easy enough, so sometimes you just reach for the heap like in MATLAB mode. Hopefully Prem and Shuhei's work lands soon enough to stack allocate small non-escaping arrays so that user's done need to think about this.
> Alignment I'd assumed, but padding the struct instead of the tuple did nothing, so probably extra work to clear a piece of an simd load. Any insight on why avx availability didn't help would be appreciated. I did verify some avx instructions were in the asm it generated, so it knew, it just didn't use.
The major differences at this point seem to come down to GCC (g++) vs LLVM and proofs of aliasing. LLVM's auto-vectorizer isn't that great, and it seems to be able to prove 2 arrays are not aliasing less reliably. For the first part, some people have just improved the loop analysis code from the Julia side (https://github.com/JuliaSIMD/LoopVectorization.jl), forcing SIMD onto LLVM can help it make the right choices. But for the second part you do need to do `@simd ivdep for ...` (or use LoopVectorization.jl) to match some C++ examples. This is hopefully one of the things that the JET.jl and other new analysis passes can help with, along with the new effects system (see https://github.com/JuliaLang/julia/pull/43852, this is a pretty huge new compiler feature in v1.8, but right now it's manually specified and will take time before things like https://github.com/JuliaLang/julia/pull/44822 land and start to make it more pervasive). When that's all together, LLVM will have more ammo for proving things more effectively (pun intended).
-
Vectorize function calls
This looks nice too. Seems to be maintained and it even has a vmap-function. What more can one ask for ;) https://github.com/JuliaSIMD/LoopVectorization.jl
-
Implementing dedispersion in Julia.
Have you checked out https://github.com/JuliaSIMD/LoopVectorization.jl ? It may be useful for your specific use case
-
We Use Julia, 10 Years Later
And the "how" behind Octavian.jl is basically LoopVectorization.jl [1], which helps make optimal use of your CPU's SIMD instructions.
Currently there can some nontrivial compilation latency with this approach, but since LV ultimately emits custom LLVM it's actually perfectly compatible with StaticCompiler.jl [2] following Mason's rewrite, so stay tuned on that front.
[1] https://github.com/JuliaSIMD/LoopVectorization.jl
[2] https://github.com/tshort/StaticCompiler.jl
-
Why Lisp? (2015)
Yes, and sorry if I also came off as combative here, it was not my intention either. I've used some Common Lisp before I got into Julia (though I never got super proficient with it) and I think it's an excellent language and it's too bad it doesn't get more attention.
I just wanted to share what I think is cool about julia from a metaprogramming point of view, which I think is actually its greatest strength.
> here is a hypothetical question that can be asked: would a julia programmer be more powerful if llvm was written in julia? i think the answer is clear that they would be
Sure, I'd agree it'd be great if LLVM was written in julia. However, I also don't think it's a very high priority because there are all sorts of ways to basically slap LLVM's hands out of the way and say "no, I'll just do this part myself."
E.g. consider LoopVectorization.jl [1] which is doing some very advanced program transformations that would normally be done at the LLVM (or lower) level. This package is written in pure Julia and is all about bypassing LLVM's pipelines and creating hyper efficient microkernels that are competitive with the handwritten assembly in BLAS systems.
To your point, yes Chris' life likely would have been easier here if LLVM was written in julia, but also he managed to create this with a lot less man-power in a lot less time than anything like it that I know of, and it's screaming fast so I don't think it was such a huge impediment for him that LLVM wasn't implemented in julia.
[1] https://github.com/JuliaSIMD/LoopVectorization.jl
-
A Project of One’s Own
He still holds a few land speed records he set with motorcycles he designed and built.
But I had no real hobbies or passions of my own, other than playing card games.
It wasn't until my twenties, after I already graduated college with degrees I wasn't interested in and my dad's health failed, that I first tried programming. A decade earlier, my dad was attending the local Linux meetings when away from his machine shop.
Programming, and especially performance optimization/loop vectorization are now my passion and consume most of my free time (https://github.com/JuliaSIMD/LoopVectorization.jl).
Hearing all the stories about people starting and getting hooked when they were 11 makes me feel like I lost a dozen years of my life. I had every opportunity, but just didn't take them. If I had children, I would worry for them.
-
When porting numpy code to Julia, is it worth it to keep the code vectorized?
Julia will often do SIMD under the hood with either a for loop or a broadcasted version, so you generally shouldn't have to worry about it. But for more advanced cases you can look at https://github.com/JuliaSIMD/LoopVectorization.jl
-
Julia 1.6 Highlights
Very often benchmarks include compilation time of julia, which might be slow. Sometimes they rightfully do so, but often it's really apples and oranges when benchmarking vs C/C++/Rust/Fortran. Julia 1.6 shows compilation time in the `@time f()` macro, but Julia programmers typically use @btime from the BenchmarkTools package to get better timings (e.g. median runtime over n function calls).
I think it's more interesting to see what people do with the language instead of focusing on microbenchmarks. There's for instance this great package https://github.com/JuliaSIMD/LoopVectorization.jl which exports a simple macro `@avx` which you can stick to loops to vectorize them in ways better than the compiler (=LLVM). It's quite remarkable you can implement this in the language as a package as opposed to having LLVM improve or the julia compiler team figure this out.
See the docs which kinda read like blog posts: https://juliasimd.github.io/LoopVectorization.jl/stable/
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.
What are some alternatives?
julia - The Julia Programming Language
cunumeric - An Aspiring Drop-In Replacement for NumPy at Scale
StaticCompiler.jl - Compiles Julia code to a standalone library (experimental)
awesome-quant - A curated list of insanely awesome libraries, packages and resources for Quants (Quantitative Finance)
cl-cuda - Cl-cuda is a library to use NVIDIA CUDA in Common Lisp programs.
cudf - cuDF - GPU DataFrame Library
julia-vim - Vim support for Julia.
Tullio.jl - ⅀
cmu-infix - Updated infix.cl of the CMU AI repository, originally written by Mark Kantrowitz
GPUCompiler.jl - Reusable compiler infrastructure for Julia GPU backends.
bel - An interpreter for Bel, Paul Graham's Lisp language
CudaPy - CudaPy is a runtime library that lets Python programmers access NVIDIA's CUDA parallel computation API.