Nvidia CEO Jensen Huang released a raft of news at the Mobile World Congress in Los Angeles, announcing a new edge supercomputing platform, a software developer kit, and fresh partnerships with Microsoft, Red Hat and Ericsson, all aimed at showcasing Nvidia’s graphics processing units (GPUs) and how these can power new 5G services and systems.
Putting out 5G services with the help of EGX
The new cloud-native Nvidia EGX Edge Supercomputing platform, is aimed at letting firms use streaming data from factory floors, manufacturing inspection lines and city streets for AI, IoT and 5G services at scale and with low latency. Early adopters for the system include Walmart, BMW, Procter & Gamble, Samsung, NTT East, and the cities of San Francisco and Las Vegas.
Building 5G RAN with Nvidia Aerial
The Nvidia Aerial software developer kit will let providers build and deliver high-performance, software-defined 5G wireless radio access networks (RAN). This will be done using a low-latency data path from Mellanox network interface cards to GPU memory, and a 5G physical layer signal-processing engine to keep data with the GPU’s memory.
Moving towards 5G networks using Red Hat and Kubernetes
With Red Hat, Nvidia puts the spotlight on software-defined 5G wireless infrastructure, running on Red Hat OpenShift, and builds on the companies’ existing relationship to accelerate the adoption of Kubernetes in enterprise data centres. The idea of the latest cooperation is to help telecom firms move towards 5G networks capable of running a range of software-defined edge workloads. Work will at the initial stage focus on 5G RAN, aimed at making AI-enabled applications more accessible at the edge. Customers will be able to use the Nvidia EGX platform and Red Hat OpenShift to more easily deploy Nvidia GPUs to push AI, data science and machine learning at the edge.
Bringing new AR, AI and gaming services to market faster with Ericsson
With Ericsson, Nvidia will build high-performing, efficient and fully virtualised 5G RAN. The partnership will combine Ericsson’s expertise in RAN technology with Nvidia’s knowledge of GPU-powered accelerated computing platform, AI and supercomputing. Nvidia said the ultimate goal is to commercialise virtualised RAN technologies that will bring new services, such as augmented reality, virtual reality and gaming, the market more quickly and with greater flexibility.
Cooperating on edge computing with Microsoft
Microsoft and Nvidia will meanwhile work on intelligent edge computing, so that industries can better manage and get insight from the data generated by retail stores, warehouses, manufacturing facilities, connected buildings, urban infrastructure and other environments. Nvidia’s Metropolis video analytics application framework, which runs on EGX, will work with Microsoft’s Azure IoT Edge, Azure Maching Learning systems and a new form factor of the Azure Data Box Edge application, powered by Nvidia T4 GPUs. Also, NVIDIA-certified off-the-shelf servers, set up to run Azure IoT Edge and ML services, are now available from over a dozen leading OEMs, including Dell, HPEe and Lenovo.
