microsoft / superbenchmark
A validation and profiling tool for AI infrastructure
☆302Updated this week
Alternatives and similar repositories for superbenchmark:
Users that are interested in superbenchmark are comparing it to the libraries listed below
- MSCCL++: A GPU-driven communication stack for scalable AI applications☆311Updated this week
- Microsoft Collective Communication Library☆343Updated last year
- A tool for bandwidth measurements on NVIDIA GPUs.☆392Updated last month
- RDMA and SHARP plugins for nccl library☆183Updated 2 months ago
- Synthesizer for optimal collective communication algorithms☆106Updated 11 months ago
- NCCL Profiling Kit☆127Updated 8 months ago
- NCCL Fast Socket is a transport layer plugin to improve NCCL collective communication performance on Google Cloud.☆116Updated last year
- ☆331Updated 11 months ago
- PArametrized Recommendation and Ai Model benchmark is a repository for development of numerous uBenchmarks as well as end to end nets for…☆132Updated last week
- NVIDIA Inference Xfer Library (NIXL)☆191Updated this week
- Microsoft Collective Communication Library☆60Updated 4 months ago
- A low-latency & high-throughput serving engine for LLMs☆327Updated last month
- oneAPI Collective Communications Library (oneCCL)☆225Updated 2 weeks ago
- ROCm Communication Collectives Library (RCCL)☆308Updated this week
- This is a plugin which lets EC2 developers use libfabric as network provider while running NCCL applications.☆167Updated last week
- Efficient and easy multi-instance LLM serving☆339Updated this week
- Fine-grained GPU sharing primitives☆141Updated 5 years ago
- NVIDIA Data Center GPU Manager (DCGM) is a project for gathering telemetry and measuring the health of NVIDIA GPUs☆480Updated last month
- AI Accelerator Benchmark focuses on evaluating AI Accelerators from a practical production perspective, including the ease of use and ver…☆229Updated this week
- NCCL Tests☆1,040Updated last week
- Kubernetes Scheduler for Deep Learning☆260Updated 2 years ago
- NVIDIA NCCL Tests for Distributed Training☆85Updated last week
- Splits single Nvidia GPU into multiple partitions with complete compute and memory isolation (wrt to performace) between the partitions☆158Updated 5 years ago
- Unified Collective Communication Library☆237Updated this week
- Dynamic Memory Management for Serving LLMs without PagedAttention☆317Updated this week
- ☆49Updated 2 weeks ago
- GPU-scheduler-for-deep-learning☆203Updated 4 years ago
- Hooked CUDA-related dynamic libraries by using automated code generation tools.☆150Updated last year
- An interference-aware scheduler for fine-grained GPU sharing☆129Updated last month
- Zero Bubble Pipeline Parallelism☆373Updated 3 weeks ago