Boost ML on Heterogeneous AI Accelerator with Ray on K8s
Talkat KubeCon + CloudNativeCon + Open Source Summit China 2023 · on Wednesday, 2023-09-27· at 16:40· 2层 会议室 2 | 2F Room 2·
summary
Describes integrating transparent backend acceleration technologies into Ray running on Kubernetes to automatically boost ML performance on heterogeneous AI accelerators. Names ML graph compilers such as TVM, Intel OpenVINO and NVIDIA TensorRT and explains using them together with Ray on k8s to provide unified software acceleration for diverse vendor accelerators without requiring code changes to user applications.
AI-written summary · about
related talks
28mScaling Multimodal Data and Reinforcement Learning with Ray
29mScaling ML Pipelines for Autonomous Driving with Ray
26mA Growing Ecosystem of Scalable ML Libraries on Ray
44mPanel: ML Platform on Ray
26mIntroduction to Distributed ML Workloads with Ray on Kubernetes
23mHeterogeneous Pipelines as the Unlock for ML at Scale
16mKeynote: An Open-Source Stack for AI Compute: Kubernetes + Ray + PyTorch + vLLM
22mRay, a Unified Distributed Framework for the Modern AI Stack