LF AI & Data Foundation - the organisation that is building an ecosystem to support open source innovation in artificial intelligence (AI) and data open source projects - today announced Kompute[1] as its latest sandbox project.
Kompute, published and open-sourced by The Institute for Ethical AI & Machine Learning[2], is a general-purpose GPU computing framework for AI and machine learning applications running across vendor graphics cards (AMD, Qualcomm, NVIDIA and friends). The Kompute framework provides a flexible interface that can be adopted by mobile, desktop, cloud and edge applications to support highly optimised GPU acceleration. The framework includes a high-level Python interface [3], which supports high-level data processing use cases, and an extensible low-level C++ interface [4], which provides high-performance device-specific optimizations [5].
The core goal of the Kompute project is to contribute to and further develop the GPU computing ecosystem in scientific and industrial applications through cross-vendor graphics card tools and capabilities. We have seen positive acceptance and adoption of Kompute by various development communities, including mobile applications[6], game development engines[7], advanced data processing use cases for edge devices and cloud computing, and we look forward to Kompute becoming a platform that empowers discussions on topics related to cross-industry applications of GPU acceleration.
The core goal of the Kompute project is to contribute to and further develop the GPU computing ecosystem in scientific and industrial applications through cross-vendor graphics card tools and capabilities. We have seen positive acceptance and adoption of Kompute by various development communities, including mobile applications[6], game development engines[7], advanced data processing use cases for edge devices and cloud computing, and we look forward to Kompute becoming a platform that empowers discussions on topics related to cross-industry applications of GPU acceleration.
LF AI & Data supports the project through a wide range of services and the first step is to join as an incubation project. Learn more about Kompute on their GitHub[8] and join the Kompute-Announce[9] and Kompute-Technical-Discuss[10] mailing lists. We also encourage you to subscribe to the monthly Kompute GPU Acceleration Conference[11], join the community and stay informed - subscription instructions are found here[12].
A warm welcome to Kompute! We look forward to the continued growth and success of this project as part of the LF AI & Data Foundation. To find out how you can host an open source project with us, please visit the LF AI & Data website [13].
References
[1] Kompute: https://github.com/KomputeProject/kompute
[2] Institute for the Ethics of Artificial Intelligence and Machine Learning: https://ethical.institute/
[3] High-level Python interfaces: https://towardsdatascience.com/beyond-cuda-gpu-accelerated-python-for-machine-learning-in-cross-vendor-graphics-cards-made-simple-6cc828a45cc3
[4] Low-level C++ interface: https://towardsdatascience.com/machine-learning-and-data-processing-in-the-gpu-with-vulkan-kompute-c9350e5e5d3a
[5] High-performance, device-specific optimisation: https://towardsdatascience.com/parallelizing-heavy-gpu-workloads-via-multi-queue-operations-50a38b15a1dc
[6] Mobile applications: https://towardsdatascience.com/gpu-accelerated-machine-learning-in-your-mobile-applications-using-the-android-ndk-vulkan-kompute-1e9da37b7617
[7] Development engine: https://towardsdatascience.com/supercharging-game-development-with-gpu-accelerated-ml-using-vulkan-kompute-the-godot-game-engine-4e75a84ea9f0
[8] GitHub: http://github.com/KomputeProject/kompute
[9] Kompute-Announce: https://lists.lfaidata.foundation/g/kompute-announce
[10] Kompute-Technical-Discuss: https://lists.lfaidata.foundation/g/Kompute-Technical-Discuss
[11] Monthly Kompute GPU Acceleration Conference: https://lists.lfaidata.foundation/g/Kompute-Technical-Discuss/viewevent?repeatid=39182&eventid=1256207&calstart=2021-09-13
[12] Here: https://wiki.lfaidata.foundation/pages/viewpage.action?pageId=46989731
[13] LF AI & Data website: https://lfaidata.foundation/proposal-and-hosting-process/