Story image

Red Hat and NVIDIA extend alliance to speed up AI

10 May 2019
Twitter

Global open source vendor Red Hat has expanded its alliance with GPU manufacturer NVIDIA.

Red Hat and NVIDIA have been partners for many years, collaborating on technologies as diverse as video drivers, heterogeneous memory management (HMM), KVM support for virtual GPUs, and Kubernetes.

Today's announcement recognises increased interest amongst enterprises in performance-sensitive workloads that can be moved from the CPU to the GPU processors. These workloads include fraud detection, image and voice-recognition, natural language processing as well as many other artificial intelligence and machine learning use cases.

Red Hat's Enterprise Linux operating system supports a bunch of NVIDIA's innovative GPU hardware, CUDA-X acceleration libraries, and NGC container registry.

The latest move sees the companies demonstrating new capabilities for running GPU-accelerated workloads across the hybrid cloud. The NGC container registry provides a rich catalogue of GPU-accelerated AI/ML and data analytics containers that can run on a foundation of Red Hat Enterprise Linux and Red Hat OpenShift Container Platform with supported NVIDIA GPUs on-premise and in the cloud.

In October 2018, Red Hat announced the certification of Red Hat Enterprise Linux on NVIDIA DGX-1 systems, opening the door for enterprises to more easily move GPU-accelerated workloads from experimental, sandbox test environments into production. Building on this work, the companies have now certified Red Hat Enterprise Linux on additional NVIDIA-powered computing platforms, including NVIDIA DGX-2 AI systems and NVIDIA T4 GPU-powered systems available from the world's leading server manufacturers.

Red Hat OpenShift Container Platform is the industry's most comprehensive enterprise Kubernetes platform: a security-focused, consistent foundation to deliver applications across the hybrid cloud, with full-stack automated operations and more streamlined developer workflows to help get to market faster. In collaboration with NVIDIA, the platform is extended to support NVIDIA GPU-accelerated systems in container environments.

"Technologies like artificial intelligence and machine learning are fundamentally changing the way we do business; however, enterprises are still facing challenges related to compatibility and support for these workloads. Through our work with NVIDIA, we are combining leading hardware and software to deliver solutions to address these challenges and help lay the foundation for data-driven innovation to thrive," says Chris Wright, chief technology officer, Red Hat.

"Today's enterprises running AI and analytics workloads at scale need world-class accelerated computing technology combined with easy-to-use, advanced software management tools. Through our expanded collaboration, NVIDIA and Red Hat are enabling businesses and organisations worldwide to more effortlessly explore and deploy AI and analytics at scale," says Chris Lamb, vice president, Computing Software, NVIDIA.

"High-performance technologies are moving at a brisk rate into enterprise data centres to accelerate product development and business operations - including financial services, ERP and sales analysis, fraud detection and cybersecurity, and machine learning-AI. The hybrid cloud solutions from Red Hat and NVIDIA are designed to make accelerated computing use easier for enterprises on-premise and in the cloud," says Steve Conway, senior vice president of research, Hyperion Research.