pytorch_dlprim
ArrayFire
pytorch_dlprim | ArrayFire | |
---|---|---|
3 | 6 | |
208 | 4,413 | |
- | 0.7% | |
5.9 | 7.1 | |
about 1 month ago | about 1 month ago | |
C++ | C++ | |
MIT License | BSD 3-clause "New" or "Revised" License |
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.
pytorch_dlprim
-
Linus Tech Tips: "China doesn't want me to have this GPU [Moore Threads MTT S80]" (Linus Tech Tips Reviews the Moore Threads MTT S80 GPU)
I know PyTorch supports OpenCL nows and you can do training with it as well. See here. Never try it myself.
-
[P] OpenCL backend for PyTorch - progress works with mainstream pytorch
I'm working on PyTorch OpenCL backend based on dlprimitives core library. It exists for a while but until now it required building custom pytorch version.
- [P] Progress with OpenCL backend for pytorch
ArrayFire
-
Learn WebGPU
Loads of people have stated why easy GPU interfaces are difficult to create, but we solve many difficult things all the time.
Ultimately I think CPUs are just satisfactory for the vast vast majority of workloads. Servers rarely come with any GPUs to speak of. The ecosystem around GPUs is unattractive. CPUs have SIMD instructions that can help. There are so many reasons not to use GPUs. By the time anyone seriously considers using GPUs they're, in my imagination, typically seriously starved for performance, and looking to control as much of the execution details as possible. GPU programmers don't want an automagic solution.
So I think the demand for easy GPU interfaces is just very weak, and therefore no effort has taken off. The amount of work needed to make it as easy to use as CPUs is massive, and the only reason anyone would even attempt to take this on is to lock you in to expensive hardware (see CUDA).
For a practical suggestion, have you taken a look at https://arrayfire.com/ ? It can run on both CUDA and OpenCL, and it has C++, Rust and Python bindings.
-
seeking C++ library for neural net inference, with cross platform GPU support
What about Arrayfire. https://github.com/arrayfire/arrayfire
-
[D] Deep Learning Framework for C++.
Low-overhead — not our goal, but Flashlight is on par with or outperforming most other ML/DL frameworks with its ArrayFire reference tensor implementation, especially on nonstandard setups where framework overhead matters
-
[D] Neural Networks using a generic GPU framework
Looking for frameworks with Julia + OpenCL I found array fire. It seems quite good, bonus points for rust bindings. I will keep looking for more, Julia completely fell off my radar.
- Windows 11 va bloquer les bidouilles qui facilitent l'emploi d'un navigateur alternatif à Edge
-
Arrayfire progressive performance decline?
Your Problem may be the lazy evaluation, see this issue: https://github.com/arrayfire/arrayfire/issues/1709
What are some alternatives?
dlprimitives - Deep Learning Primitives and Mini-Framework for OpenCL
Thrust - [ARCHIVED] The C++ parallel algorithms library. See https://github.com/NVIDIA/cccl
oneDNN - oneAPI Deep Neural Network Library (oneDNN)
Boost.Compute - A C++ GPU Computing Library for OpenCL
mace - MACE is a deep learning inference framework optimized for mobile heterogeneous computing platforms.
VexCL - VexCL is a C++ vector expression template library for OpenCL/CUDA/OpenMP
Pytorch - Tensors and Dynamic neural networks in Python with strong GPU acceleration
FluidX3D - The fastest and most memory efficient lattice Boltzmann CFD software, running on all GPUs via OpenCL.
CUB - THIS REPOSITORY HAS MOVED TO github.com/nvidia/cub, WHICH IS AUTOMATICALLY MIRRORED HERE.
Taskflow - A General-purpose Parallel and Heterogeneous Task Programming System
moderngpu - Patterns and behaviors for GPU computing