Verizon, Nvidia launch 5G-powered AI service for enterprises

As artificial intelligence (AI) is now totally established in the fabric of networks, accelerating across applications such as collaboration and connected vehicle infrastructures, leading communications provider Verizon has established a partnership with AI chip leader Nvidia to enable a range of AI applications to run over a 5G private network with private mobile edge computing (MEC).

The result, said the two companies, will be to develop a “groundbreaking” offering that merges AI with the speed of private 5G networks to no less than “transform” a range of industries and drive significant economic growth. Initial demonstrations of the service will begin in February 2025.

Verizon and Nvidia said the need for networks to manage AI workloads has never been greater, citing PwC’s Global artificial intelligence study, which reports that 75% of business executives consider AI a business advantage, and are either already using it or planning to do so.

In addition, they noted the McKinsey global survey on AI indicating that 50% of companies have adopted AI in at least one business function, with 71% planning to expand AI use in the near future. These insights, they said, underscore the importance of mobile networks being able to support AI workloads in the future.

In practice, the partnership will combine Verizon’s 5G Private Network with Enterprise AI and private MEC together with the Nvidia AI Enterprise software platform and Nvidia NIM microservices to deliver real-time AI services on-premise for enterprise customers.

Key benefits of the service are said to include the ultra-low latency required for real-time AI processing at the edge, critical for applications such as robotics, and augmented reality; dedicated bandwidth and enhanced reliability to ensure optimal performance for data-intensive AI workloads; the necessary improved security and control over data and applications for businesses with sensitive AI workloads; quick decision-making with local processing and inferencing through AI processing at the Edge; and tailoring to meet the specific needs of different industries and use cases.

The AI-powered private 5G platform stack is designed to be plug and play, helping third-party developers innovate with speed, while also accommodating future evolutions in AI computing and AI and connectivity applications. It can support multi-tenancy for multiple use cases or customers, and is modular to scale as needed for a bespoke service. It can also provide services remotely via portable private network offerings or on a customer’s premise with a permanent private network on-site.

Compute-intensive apps the stack is built to handle include generative AI large language models and vision language models, video streaming, broadcast management, computer vision, augmented, virtual and extended reality, autonomous mobile robot and automated guided vehicle, and internet of things.

“Generative AI, with its predictive capabilities, is poised to be a critical component of digital transformation and future business growth in almost all industries,” said Srini Kalapala, senior vice-president of technology and product development at Verizon. “In today’s rapidly evolving digital landscape, staying ahead of the curve is essential for businesses to thrive.”

Ronnie Vasishta, senior vice-president of telecom at Nvidia, added: “Enterprises everywhere are racing to integrate AI solutions that bring new value to their employees, partners and customers, and can also help them operate with extreme efficiency.

“Verizon’s integration of Nvidia’s full-stack AI platform into its new solution for running AI workloads on private 5G networks is a big step forward in helping enterprises of all sizes reach their business objectives faster with AI.”