ParCoreLab / ComScribeLinks
ComScribe is a tool to identify communication among all GPU-GPU and CPU-GPU pairs in a single-node multi-GPU system.
☆27Updated 2 years ago
Alternatives and similar repositories for ComScribe
Users that are interested in ComScribe are comparing it to the libraries listed below
Sorting:
- [ACM EuroSys 2023] Fast and Efficient Model Serving Using Multi-GPUs with Direct-Host-Access☆56Updated 5 months ago
- ☆38Updated 5 years ago
- ☆44Updated 4 years ago
- NCCL Profiling Kit☆150Updated last year
- Thunder Research Group's Collective Communication Library☆47Updated 6 months ago
- Artifact of OSDI '24 paper, ”Llumnix: Dynamic Scheduling for Large Language Model Serving“☆64Updated last year
- Microsoft Collective Communication Library☆66Updated last year
- SOTA Learning-augmented Systems☆37Updated 3 years ago
- Paella: Low-latency Model Serving with Virtualized GPU Scheduling☆66Updated last year
- An experimental parallel training platform☆56Updated last year
- ☆56Updated 5 years ago
- Fine-grained GPU sharing primitives☆147Updated 6 months ago
- ☆47Updated last year
- ☆84Updated 3 years ago
- A GPU-accelerated DNN inference serving system that supports instant kernel preemption and biased concurrent execution in GPU scheduling.☆44Updated 3 years ago
- NCCL Fast Socket is a transport layer plugin to improve NCCL collective communication performance on Google Cloud.☆122Updated 2 years ago
- ☆26Updated 11 months ago
- DISB is a new DNN inference serving benchmark with diverse workloads and models, as well as real-world traces.☆58Updated last year
- ☆17Updated 3 years ago
- PipeSwitch: Fast Pipelined Context Switching for Deep Learning Applications☆127Updated 3 years ago
- Bamboo is a system for running large pipeline-parallel DNNs affordably, reliably, and efficiently using spot instances.☆55Updated 3 years ago
- High performance RDMA-based distributed feature collection component for training GNN model on EXTREMELY large graph☆56Updated 3 years ago
- ☆24Updated 2 years ago
- ☆25Updated 2 years ago
- REEF is a GPU-accelerated DNN inference serving system that enables instant kernel preemption and biased concurrent execution in GPU sche…☆104Updated 3 years ago
- RDMA and SHARP plugins for nccl library☆221Updated 2 weeks ago
- ☆16Updated 9 months ago
- Multi-GPU communication profiler and visualizer☆37Updated last year
- Analysis for the traces from byteprofile☆32Updated 2 years ago
- GPUDirect Async support for IB Verbs☆135Updated 3 years ago