paperg / NCCL_GPLinks
Separate from hardware and used to learn some NCCL mechanisms
☆23Updated last year
Alternatives and similar repositories for NCCL_GP
Users that are interested in NCCL_GP are comparing it to the libraries listed below
Sorting:
- ☆214Updated last week
- ☆205Updated 2 years ago
- example code for using DC QP for providing RDMA READ and WRITE operations to remote GPU memory☆144Updated last year
- RDMA and SHARP plugins for nccl library☆208Updated 3 weeks ago
- NCCL Profiling Kit☆145Updated last year
- ☆30Updated last year
- ☆67Updated 4 months ago
- ☆42Updated 11 months ago
- TACCL: Guiding Collective Algorithm Synthesis using Communication Sketches☆75Updated 2 years ago
- DeepSeek-V3/R1 inference performance simulator☆170Updated 6 months ago
- ☆46Updated 9 months ago
- ☆85Updated 6 months ago
- Repository for MLCommons Chakra schema and tools☆127Updated last week
- ☆18Updated 7 months ago
- Hooked CUDA-related dynamic libraries by using automated code generation tools.☆167Updated last year
- Microsoft Collective Communication Library☆360Updated 2 years ago
- AI Accelerator Benchmark focuses on evaluating AI Accelerators from a practical production perspective, including the ease of use and ver…☆266Updated last month
- This is an RDMA program written in Python, based on the Pyverbs provided by the rdma-core(https://github.com/linux-rdma/rdma-core) reposi…☆33Updated 3 years ago
- Paella: Low-latency Model Serving with Virtualized GPU Scheduling☆62Updated last year
- Synthesizer for optimal collective communication algorithms☆117Updated last year
- ☆37Updated 3 years ago
- ASTRA-sim2.0: Modeling Hierarchical Networks and Disaggregated Systems for Large-model Training at Scale☆441Updated last month
- Artifact from "Hardware Compute Partitioning on NVIDIA GPUs". THIS IS A FORK OF BAKITAS REPO☆34Updated last year
- Injecting Adrenaline into LLM Serving: Boosting Resource Utilization and Throughput via Attention Disaggregation☆37Updated 2 weeks ago
- A prefill & decode disaggregated LLM serving framework with shared GPU memory and fine-grained compute isolation.☆111Updated 4 months ago
- NCCL Fast Socket is a transport layer plugin to improve NCCL collective communication performance on Google Cloud.☆120Updated last year
- ☆154Updated last year
- Artifacts for our NSDI'23 paper TGS☆86Updated last year
- Ultra and Unified CCL☆567Updated this week
- Thunder Research Group's Collective Communication Library☆42Updated 2 months ago