NVIDIA FabricManager Go Package

Build NVIDIA® CUDA™ code for OpenCL™ 1.2 devices

Documentation of NVIDIA chip/hardware interfaces

Go heap object reference analysis tool

Go RPC framework with high-performance and strong-extensibility for building micro-services.

A high-performance non-blocking I/O networking framework focusing on RPC scenarios.

A do everything Redfish, KVM, GUI, and DBus webserver for OpenBMC

A do everything Redfish, KVM, GUI, and DBus webserver for OpenBMC

KAI Scheduler is an open source Kubernetes Native scheduler for AI workloads at large scale

🌸 Run LLMs at home, BitTorrent-style. Fine-tuning and inference up to 10x faster than offloading

GPU Admin Tools. Includes Confidential Computing controls for H100, and other functionality

Ancillary open source software to support confidential computing on NVIDIA GPUs

Resource scheduling and cluster management for AI

Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels

Tile-Based Runtime for Ultra-Low-Latency LLM Inference

Tool to collect debug logs from NVIDIA server components, in band and out-of-band.

a fast and customizable CUDA int4 tensor core gemm

NCCL Tests

A command line utility to manage the configuration of a system's high performance network interfaces for RoCE deployments

Translation of C++ Core Guidelines [https://github.com/isocpp/CppCoreGuidelines] into Simplified Chinese.

slime is an LLM post-training framework for RL Scaling.

A NCCL extension library, designed to efficiently offload GPU memory allocated by the NCCL communication library.

LMCache on Ascend

torchcomms: a modern PyTorch communications API

Project infrastructure administrative tools

DLRover: An Automatic Distributed Deep Learning System

Perplexity GPU Kernels

Source code examples from the Parallel Forall Blog