Accelerate AI with NVIDIA and Red Hat

Speed up delivery of AI-powered intelligent apps across datacenter, edge, and public clouds.

Jump to section

Red Hat® and NVIDIA® are collaborating to enable the latest technological innovations. We share a vision of an IT future fueled by cloud-native, open source technologies where our customers can use familiar infrastructure on many footprints, from virtualized and private cloud deployments in corporate datacenters, to massive-scale services on public clouds, and all the way to the edge.

NVIDIA AI Enterprise on Red Hat OpenShift

NVIDIA AI Enterprise is an end-to-end, cloud-native suite of AI and data analytics software, optimized, certified, and supported by NVIDIA to run on Red Hat OpenShift, the industry leading Kubernetes Platform, with NVIDIA-Certified Systems. It includes key technologies from NVIDIA and Red Hat to securely deploy, manage, and scale AI workloads consistently across the hybrid cloud.

Read the solution brief to learn more about NVIDIA AI Enterprise on Red Hat OpenShift.

Green and gray chart of Data scientist and IT Admin workflow

Accelerated AI infrastructure for datacenters

NVIDIA and Red Hat provide accelerated AI infrastructure targeted for on-premise deployments.

Based on Red Hat OpenShift, this joint solution is a platform for accelerating ML/DL modeling for data scientists and delivers intelligent, AI-powered applications.

Organizations can use NVIDIA’s NGC™ software hub to access GPU-optimized DL and ML applications and pre-trained AI models to take advantage of massive parallel processing capabilities.Leading server providers offer NVIDIA-Certified Systems powered by NVIDIA GPUs, which are validated to run accelerated applications with optimal performance.

IT manageability for purpose-built AI infrastructure

NVIDIA and Red Hat bring the enterprise IT benefits of Red Hat Enterprise Linux® and Red Hat OpenShift to NVIDIA DGX™ systems, NVIDIA’s infrastructure platform is purpose-built for the unique demands of AI. Organizations can deploy these universal systems for all AI workloads within the familiar management paradigm of Red Hat Enterprise Linux—employing IT-developed methods and optimized AI software from NVIDIA. As a NVIDIA DGX-Ready Software program member, Red Hat OpenShift is certified for use with DGX systems.

Nvidia hardware

NVIDIA GPU Operator for Red Hat OpenShift 

NVIDIA GPU Operator for Red Hat OpenShift helps simplify and accelerate compute-intensive ML/DL modeling tasks for data scientists, and inferencing tasks across datacenters, edge, and public clouds. The GPU Operator enables Red Hat OpenShift to schedule workloads that require NVIDIA GPUs as easily scheduling CPU or memory for traditional workloads. Typical use cases that can benefit from GPU acceleration include image and speech recognition, conversational AI, and visual search.

The joint solution is compatible with NVIDIA-Certified Systems  and Red Hat Enterprise Linux® certified servers.

Graphic displaying operator to management flow

Accelerated Networking for Modern Datacenters

Hardware & software must work hand in hand to build modern datacenters. NVIDIA and Red Hat have a long history of working together to provide accelerated networking solutions that create robust, scalable, and secure infrastructure for modern workloads.

Through mutual collaboration, we can help you speed the deployment of optimized software-defined infrastructure, accelerate the development of applications to unlock improved operational efficiency and competitive differentiation.

Red Hat and NVIDIA also recognize the need to deliver standardized and accessible infrastructure for powering enterprise digital transformation. We work closely to bring networking hardware innovations to cloud-native computing through Red Hat OpenShift and cloud networking through OpenStack. By utilizing NVIDIA technologies including GPUDirect RDMA, ASAP2 (accelerated switching and packet processing), and the Morpheus AI cybersecurity framework, we enable enterprises to securely deliver the next wave of applications.

Nvidia 3d learning inference graphic showing 4 quadrants

Record breaking STAC-A2 Market Risk benchmark

Red Hat and NVIDIA collaborated to create the latest STAC-A2 entrant, and STAC performed an independent, third-party audit. These are the first public STAC-A2 results using Red Hat OpenShift. Compared to all other publicly reported results to-date, this solution based on Red Hat OpenShift and NVIDIA DGX A100 set several new records for performance metrics and energy efficiency:

  • highest energy efficiency
  • highest space efficiency2
  • highest throughput3
  • fastest warm & cold start times in the large Greeks benchmark4
  • fastest warm time in the baseline Greeks benchmark5
  • highest maximum paths6 
  • highest maximum assets7

Below is the high-level architectural diagram of the benchmark configuration that was used.

Stac-A2 configuration

Webinar: AI-enabled data science platform for financial services

With the promise of AI, financial institutions are racing to AI-enable hundreds of applications to better serve customers.

Customer case study

Israel's Ministry of Defense operationalizes ML on OpenShift with NVIDIA DGX

Press release

Red Hat Expands Alliance with NVIDIA to Accelerate and Scale AI/ML Workloads cross Hybrid Cloud Environments