Helping telecoms leverage edge, artificial intelligence (AI), and perhaps least surprisingly 5G to drive new service revenues, empower smart cities and factories, and reign in cyberthreats, was an ongoing theme throughout last week's MWC conference in Barcelona.

During a panel hosted by Palo Alto Networks, the security vendor made its case for how secure access service edge (SASE) and 5G network slicing could be combined not only to offer customers more robust and granular service level agreements, but also provide better protection against cybercriminals to boot.

Meanwhile, chip giant Nvidia announced a new lab in collaboration with Google Cloud to jumpstart the development of AI workloads running at the 5G edge, and unveiled its 5G-on-AI server platform.

“We believe every industry will be transformed in the next 10 years,” said Ronnie Vasishta, SVP of telecoms at Nvidia. “This is because the forces of AI and 5G connectivity are combining with the digital automation to drive the fourth industrial revolution.”

And while SASE and GPU-accelerated AI may not be directly connected, telecoms stand to benefit from both, especially as 5G networks become more mature and technologies like network slicing become commonplace.

The world is changing and so is the threat landscape, said Keith O’Brien, CTO of worldwide service providers at Palo Alto Networks, during the event. Post-pandemic, enterprise demands have changed, hybrid work models are quickly becoming the norm, and because of this security perimeters have grown exponentially, he added.

According to Palo Alto Networks, service providers need to rethink how they deliver networking and security to address these challenges.

Carriers are no strangers to SD-WAN or managed security, and they often bundled them alongside other services like voice-over-IP. However, these services have largely centered around branch connectivity and on-premises security appliances like firewalls.

Coined by Gartner in 2019, SASE melds elements of SD-WAN, managed security, and edge compute into a single, cloud-delivered service that can be consumed from anywhere.

This flexibility, combined with a low barrier to entry and subscription model, helped to catapult enterprise adoption of SASE in the early days of the pandemic. Communications service providers, however, are only now beginning to warm up to the architecture.

AT&T and Verizon were among the first U.S. carriers to announce managed SASE services earlier this year. AT&T’s is powered by Fortinet and Palo Alto Networks, while Verizon is using a combination of Versa Networks and Zscaler.

Combined with 5G network slicing, carriers will soon be able to provide private cellular connections to customers by virtually segregating traffic from the rest of the network. In addition to being private and less prone to congestion, network slicing will enable service providers to offer better security and unique SLAs, explained Sree Koratala, VP of mobility security product management at Palo Alto Networks.

For now, 5G network slicing remains a relatively nascent technology, which requires a 5G network core — something few operators can claim.

Speaking of 5G, Nvidia’s new AI-on-5G innovation lab, announced with Google Cloud, promises to speed the deployment of AI workloads on 5G networks.

The lab combines Google’s Anthos hybrid-cloud platform with Nvidia-certified hardware and software. The goal is to provide a consistent platform for developers as they build out the services and applications at the 5G edge required to make things like smart cities and factories a reality.

“In this lab, industrial companies, system integrators, and network operators will be able to develop and test their AI-on-5G enterprise applications on Google Anthos using Nvidia AI infrastructure,” Vasishta said.

5G networks are well suited to provide the secure, low-latency connectivity required by AI workloads at scale, said Soma Velayutham, industry GM of AI and accelerated computing at Nvidia, during a virtual event last week.

One application of AI running at the 5G edge could be for real-time traffic prioritization, he explained, describing a scenario in which traffic cameras connected over 5G to an AI platform running at the edge could automatically detect a fire truck or emergency vehicle and adjust traffic patterns accordingly.

Another example would be AI-controlled assembly and inspection. “Imagine one arm placing the windscreen, another robot sealing it, and another robot inspecting the seal all in real time,” he said. “These collaborative robots need to be able to understand each other and coordinate amongst each other as the car is going through the manufacturing process. This is the promise of AI on 5G.”

Nvidia’s AI-on-5G hardware serves as both the 5G base station and an edge data center for running AI applications. The platform is built around three components: a host CPU, a GPU that handles layer-1 virtual radio access network (RAN) processes and AI workloads, and a BlueFeild-2 DPU, which serves as a standardized open RAN interface and offloads 5G user plane functions.

Nvidia claims its AI-on-5G platform will enable the creation of high-performance 5G RAN and AI applications to manage emerging use cases like robotic manufacturing, autonomous vehicles, drones, and surveillance.

“We have now brought the power of AI cloud to the 5G connected enterprise,” Vasishta said. “This brings tremendous untapped monetization opportunities to operators that have already spent billions of dollars on acquiring spectrum.”