RAFT contains fundamental widely-used algorithms and primitives for machine learning and information retrieval. The algorithms are CUDA-accelerated and form building blocks for more easily writing high performance applications.

FlashInfer: Kernel Library for LLM Serving

A faster implementation of OpenCV-CUDA that uses OpenCV objects, and more!

cuGraph - RAPIDS Graph Analytics Library

cuVS - a library for vector search and clustering on the GPU

Graphics Processing Units Molecular Dynamics

A plugin to use Nvidia GPU in PySCF package

Code repository for the massively-parallel framework for maximum-likelihood fits, implemented in CUDA/OpenMP

Maximum clique solver running on GPUs

HierarchicalKV is a part of NVIDIA Merlin and provides hierarchical key-value storage to meet RecSys requirements. The key capability of HierarchicalKV is to store key-value feature-embeddings on high-bandwidth memory (HBM) of GPUs and in host memory. It also can be used as a generic key-value storage.

Tool for get information about NVIDIA GPUs

Preconditioned conjugate gradient method with ILU preconditioner implemented in CUDA

A CUDA based sieve for numbers of the form k*b^n-1. This project is heavily based on SR2Sieve.

Enhanced the computation runtime for (C = A⊗B^T ) and (AB + CD^T ) by effectively leveraging Memory Coalescing and Shared Memory optimization techniques, while working with a 1024x1024 sized matrix.

Designed efficient room reservation system for facility rooms with flexible slot booking (1-24 time slots), leveraged GPU parallel processing techniques to concurrently process a large number of user requests.

Implemented activation rules and a depth-first hierarchy strategy, to efficiently optimize activation point requirements for each node within a large-scale graph with 10 Million vertices and 100 Million edges.

Numerical model for simulating shallow water hydrodynamics on the GPU using an Adaptive Mesh Refinment type grid. The model was designed with the goal of simulating inundation (River, Storm surge or tsunami). The model uses a Block Uniform Quadtree approach that runs on the GPU but the adaptive/multi-resolution/AMR is being implemented and not y…

CUDA based Pathtracing Offline and Realtime Renderer

Spacemesh-cuda is a library for plot acceleration using CUDA-enabled GPUs.

High performance CUDA/Python library for computing quantum chemistry density-based descriptors for larger systems using GPUs.