File and Object Storage

Its a containerized world - AI with IBM Spectrum Scale and NVIDIA

By David Wohlford posted Thu October 01, 2020 08:20 AM

As containerized environments grow and as more customers move to the cloud, storage must adapt. Data persistence and performance shoot to the top of “container data resource” needs. To stay ahead, IBM supports the CSI open initiative for IBM Spectrum Scale.

But what about AI and all the high performance applications for high performance computing?  The growth in this area is starting to expand quickly as development loves the flexibility and hybrid cloud portability of containers and Kubernetes platforms like Red Hat OpenShift.    Bringing high performance computing data to Red Hat OpenShift is now made easier with recommendations and configuration examples.  IBM Spectrum scale is the leading platform for high performance computing environments and is becoming the preferred platform for AI and high performance workloads on Red Hat OpenShift.   This is a great place to start and understand how IBM Spectrum Scale brings parallel data access to NVIDIA workloads that are running on Red Hat OpenShift.   IBM Spectrum IBM has a number of successes and best practices that have been put together in the latest red book

Deployment and Usage Guide for Running AI Workloads on Red Hat OpenShift and NVIDIA DGX Systems with IBM Spectrum Scale

IBM Spectrum Scale (ESS 3000) and NVIDIA with Red Hat OpenShift