search

open-heterogeneous-computing-framework / conference

OHCF related conference CFPs
Apache License 2.0
4 stars 2 forks source link

A AI accelerator HDDL in Edge Computing #4

Closed lvmxh closed 5 years ago

lvmxh commented 5 years ago

Abstract Intel today introduced the Movidius Myriad X Vision processing Unit (VPU) which Intel is calling the first vision processing system-on-a-chip (SoC) with a dedicated neural compute engine to accelerate deep neural network inferencing at the network edge. Intel also Unleashes a HDDL PCI card which integrates several Myriad VPUs for Edge Computing for AR/VR and AI scenarios. In this talk, we will first introduce some details of Movidius VPU and HDDL, how it is managed in Edge cloud. Also will give a demo to mange the HDDL and to use it for an AI inference.

Related Upstream Open Source Projects OpenStack Nova, OpenStack Cyborg, Kubernetes

Category Accelerator Hardware Cloud Platform Management Support For Accelerators Applications With Accelerator Support (AI, 5G, Edge/IoT, HPC, Quantum Computing, Neuromorphic Computing, etc) Speaker Bio (no more than 2 people) Shaohe Feng, Senior Software Engineers, INTEL Got involved in OpenStack for 5+ years and Kubernetes for 2+ years. Mainly focus on accelerator management. Interests include AI technology, hardware acceleration, network and so on.

Speaker email ShaoHe Feng(shaohe.feng@intel.com)

hannibalhuang commented 5 years ago

Move to part of the unconference discussion