A high-performance distributed file system designed to address the challenges of AI training and inference workloads.

Expert Parallelism Load Balancer

Analyze computation-communication overlap in V3/R1.

A bidirectional pipeline parallelism algorithm for computation-communication overlap in V3/R1 training.

DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling

DeepEP: an efficient expert-parallel communication library

FlashMLA: Efficient MLA decoding kernels

Production-tested AI infrastructure tools for efficient AGI development and community-driven innovation

MoBA: Mixture of Block Attention for Long-Context LLMs

NVIDIA Linux open GPU with P2P support

Doing simple retrieval from LLM models at various context lengths to measure accuracy

Reexamining Direct Cache Access to Optimize I/O Intensive Applications for Multi-hundred-gigabit Networks

A fast GPU memory copy library based on NVIDIA GPUDirect RDMA technology

A Flexible Framework for Experiencing Cutting-edge LLM Inference Optimizations

Mooncake is the serving platform for Kimi, a leading LLM service provided by Moonshot AI.

how to optimize some algorithm in cuda.

CUDA Templates for Linear Algebra Subroutines

Large Language Model Text Generation Inference

A high-throughput and memory-efficient inference and serving engine for LLMs

Transformer related optimization, including BERT, GPT

Inference code for Llama models

Intel® Performance Counter Monitor (Intel® PCM)

Grasper: A High Performance Distributed System for OLAP on Property Graphs.

A solver for subgraph isomorphism problems, based upon a series of papers by subsets of McCreesh, Prosser, and Trimble.

CP 2015 subgraph isomorphism experiments, data and paper

Graph Pattern Mining

PLCT实验室的公开演讲，或者决定公开的组内报告

Open-source graph database, tuned for dynamic analytics environments. Easy to adopt, scale and own.