Showing 1-20 of 145 projects
Python SDK and Proxy Server for calling 100+ LLM APIs with cost tracking and guardrails
Build and run Docker containers leveraging NVIDIA GPUs
A scalable generative AI framework for researchers and developers
Open-source NVIDIA Linux GPU kernel modules for developers working with NVIDIA hardware.
A GameStream client for PCs that supports Windows, Mac, Linux, and Steam Link devices.
CUDA on non-NVIDIA GPUs, a Rust library for utilizing CUDA on a variety of GPU architectures.
TensorRT LLM provides a Python API and optimizations to efficiently run large language models on NVIDIA GPUs.
NVIDIA TensorRT is an SDK for high-performance deep learning inference on NVIDIA GPUs.
A lightweight, open-source control tool for ASUS laptops and ROG Ally with performance, fan, GPU, battery, and RGB management.
GPU and accelerator process monitoring tool for various hardware platforms and operating systems.
A high-performance linear algebra library for GPU-accelerated deep learning and other applications.
NVIDIA CUDA samples that demonstrate features of the CUDA Toolkit for GPU-accelerated development.
A guide to deploying deep-learning inference networks and computer vision primitives with NVIDIA Jetson hardware and TensorRT.
NVIDIA's new open-source repository collection for the Cosmos platform, focused on AI tools and workflows.
OptiScaler bridges upscaling/frame gen across GPUs, supporting DLSS2+, XeSS, FSR2+ and more.
An interactive NVIDIA-GPU process viewer and management tool for developers working with GPU-accelerated applications.
This is a Windows tool to inspect and customize NVIDIA graphics driver profiles.
An open-source toolkit for easily adding programmable guardrails to LLM-based conversational systems.
A high-performance machine learning library for CUDA-enabled GPUs, accelerating data science workflows.
A Python library for serving large language models (LLMs) with high performance, including GPU acceleration and distributed inference.
Get weekly updates on trending AI coding tools and projects.