What does HackerNews think of taichi?
Productive & portable high-performance programming in Python.
Language:
C++
Look at Taichi at Github. This library for Python seems not very popular and unaware. Maybe, because it is a Chinese development, but Taichi is simple and compiles directly down to kernels on CUDA, GPU, Metal, Vulkan and has batteries included. Beats the fastest Mojo implementation of the Mandelbrot set about 260 times faster.
https://github.com/taichi-dev/taichi
Looks like what taichi(https://github.com/taichi-dev/taichi) is doing, does this support CUDA yet?
additionally how does it compare to numba the compiler for python?
looks like python's performance on ML and AI field will only get stronger.
I would love to see the OP benchmarks with Taichi applied: https://github.com/taichi-dev/taichi
You should check out taichi: https://github.com/taichi-dev/taichi They have a ton of great demos for doing physics but check out this example in particular for something related to your project: https://github.com/taichi-dev/quantaichi#game-of-life-gol. (Taichi also makes it super easy to write things for the GPU and the kernels are differentiable :).)
Thanks for the kind words, I think that crown is for MKL-DNN / OneDNN though, for now ;), and in the deep learning domain.
I also expect Halide[1], Taichi[2], DaCe[3] and Tiramisu[4] to be able to more easily reach those levels of performance hence why I want to add a compiler to Laser (implemented at compile-time via Nim macros) but prerequisite is excellent and composable multithreading support which OpenMP didn't provide.
In statistics / data analysis, my PCA though (Principal Component Analysis) beats all implementations I've seen from Scipy, Facebook and R.
[2]: https://github.com/taichi-dev/taichi