Nvidia and Google Cloud to Create Industry’s First AI-on-5G Lab to Accelerate AI Development

On June 28, NVIDIA announced that it will cooperate with Google Cloud (Google Cloud) to establish the industry’s first AI-on-5G Innovation Lab (Innovation Lab) in the second half of 2021, enabling network infrastructure developers and AI ( artificial intelligence) software partners are able to develop, test and adopt solutions to accelerate the creation of smart cities, smart factories and other advanced 5G and AI applications.

The lab will provide enterprises with access to Google Cloud’s Anthos platform and Nvidia’s accelerated computing hardware and software platforms to leverage data and AI to improve business performance, improve operational efficiency, and optimize security and reliability.

Ronnie Vasishta, Senior Vice President of Telecom at NVIDIA, said: “The global telecom industry is undergoing a sweeping transformation as 5G speeds and low latency continue to create unprecedented opportunities to deliver AI-on-5G at the enterprise forefront. This partnership with Google Cloud , we will help network operators and infrastructure developers to conceive and create new profit centers based on AI and machine learning.”

Expanding the AI-on-5G Ecosystem

It is understood that as early as April this year, NVIDIA announced that it has cooperated with Google Cloud, Fujitsu, Mavenir, Radisys and Wind River to develop solutions based on NVIDIA’s AI-on-5G platform. Google Cloud extends the Anthos application platform to the network edge, enabling telecom service providers and enterprises to rapidly deliver new services and applications at the 5G edge.

“With the Anthos Application Platform, Google Cloud enables service providers and enterprises in many vertical markets to seamlessly connect over 5G at the network edge,” said Shailesh Shukla, Vice President and General Manager, Networking, Google Cloud. “We are excited to expand our partnership with NVIDIA, Provides a joint innovation lab that provides enterprises, CSPs and partners with the infrastructure they need to develop, test and deploy their edge AI applications.”

Anthos provides a unified platform for application deployment, with a service-centric view of each environment. The platform enables customers to build and deploy enterprise-grade containerized applications faster using managed Kubernetes (the open source container orchestration system) in the cloud, on-premises and at the network edge.

Anthos supports NVIDIA GPU-accelerated servers for a consistent scheduling and operational experience across deployments, while reducing overhead and increasing developer productivity. The platform also secures application and software supply chains and provides an outcome-focused approach to managing application policies across environments, a key capability for network operators and enterprises to deliver AI on 5G.

NVIDIA hardware and software development

NVIDIA AI-on-5G provides enterprises, mobile network operators and cloud service providers with a single converged platform for handling 5G and edge AI computing. The platform will use the NVIDIA Aerial™ software development kit on the NVIDIA BlueField™-2 A100, a fusion card that combines a GPU and DPU with NVIDIA’s “5T for 5G”, enabling the creation of high-performance 5G RAN and AI applications programs to manage precision-manufactured robots, automated guided vehicles, drones, wireless cameras, self-checkout lanes and hundreds of other transformational projects.

The NVIDIA Aerial SDK, combined with NVIDIA Metropolis, an AI framework for smart cities, NVIDIA Isaac® for automated machines, and NVIDIA Clara® for healthcare, forms part of the AI-on-5G ecosystem and can be deployed in a single NVIDIA-Certified System? certified system.

In addition, NVIDIA also announced that its next-generation Aerial A100 AI-on-5G computing platform will integrate 16 Arm-based CPU cores in the NVIDIA BlueField-3 A100 to form a single fusion card. This card delivers enterprise edge AI applications via cloud-native 5G vRAN, with improved performance per watt and faster time to deployment. These solutions will be used for testing and validation in AI-on-5G labs.

NVIDIA AI-on-5G Data Center – From Edge to Cloud

Software-defined RAN is critical to building a modern 5G infrastructure capable of running a range of applications on a common platform. NVIDIA Aerial provides elastic service with changes in network traffic throughout the day, and can provide services flexibly according to customer needs, so as to achieve optimal utilization.

Through the NVIDIA EGX™ platform from NVIDIA-Certified Systems, OEMs can bring AI computing power to the edge where data is created. Expanding on the platform, server makers can also pair Nvidia GPUs and DPUs to build hyperconverged edge data centers.

Source: Gasgoo

Author: Liu Liting