TopOpenSourceAI
zh
← Back to scenarios

GPU Compute Frameworks

Low-level frameworks and libraries for writing and running portable GPU compute kernels for data processing, simulation, and machine-learning acceleration across vendors.

7 projects

See methodology for ranking rules; order uses public GitHub metrics within this scenario.

1–7 of 7

Rank Project Stars Forks
1 kompute

General purpose GPU compute framework built on Vulkan to support 1000s of cross vendor graphics cards (AMD, Qualcomm, NVIDIA & friends). Blazing fast, mobile-enabled, asynchronous and optimized for advanced GPU data processing usecases. Backed by the Linux Foundation.

2.6K 198
2 lupine

LUPINE is a GPU over IP bridge allowing GPUs on remote machines to be attached to CPU-only machines.

2.4K 134
3 dstack

Vendor-agnostic orchestration for training, inference and agentic workloads across NVIDIA, AMD, TPU, and Tenstorrent on clouds, Kubernetes, and bare metal.

2.2K 252
4 MetalPetal

A GPU accelerated image and video processing framework built on Metal.

2.2K 288
5 tt-metal

:metal: TT-NN operator library, and TT-Metalium low level kernel programming model.

1.6K 616
6 blazingsql

BlazingSQL is a lightweight, GPU accelerated, SQL engine for Python. Built on RAPIDS cuDF.

2.0K 183
7 uccl

UCCL is an efficient communication library for GPUs, covering collectives, P2P (e.g., KV cache transfer, RL weight transfer), and EP (e.g., GPU-driven)

1.5K 171