Opanga Networks added integration with Nvidia that will allow its Rain containerized runtime to operate directly on the chip giant’s BlueField-3 digital processing unit (DPU), a move Opanga said accelerates use case support and provides a base for AI-driven autonomous networks.

The BlueField-3 integration allows the Rain platform to tap in-line, real-time customer experience insight without impacting host CPU or memory resources; tap that in-line insight for operational and customer-facing use cases; and more easily surface that data for management.

Opanga noted that the BlueField-3 platform provides the processing capacity and hardware offload needed to port real-time flow detection onto the DPU. These were previously done using network interface cards (NICs) tied to dedicated compute infrastructure.

“By executing intelligence directly on BlueField-3, Opanga reduces host CPU utilization and memory consumption while eliminating the need for additional analytics servers,” the firm explained. “Service providers can deploy advanced telemetry, scoring, and observability using infrastructure already present in their network and OSS/BSS environments.”

Ben Hadorn, VP of customer engineering and operations at Opanga, noted this ability is important as operators increasingly look toward autonomous networks.

“Autonomous networks require more than automation – they require real-time, customer-validated intelligence operating directly in the data path,” Ben Hadorn, VP of customer engineering and operations at Opanga, explained in a statement. “By running natively on Nvidia BlueField-3, we’re embedding customer experience ground truth into the infrastructure itself. That changes how operators deploy AI, validate intent, and progress toward true closed-loop operations.”

Autonomous networks were a big part of this week’s Mobile World Congress event in Barcelona, with vendors and operators repeatedly touting benefits of using AI to help run increasingly complex network operations.

Automation, T-Mobile US, and Cisco

The integration also furthers Opanga’s angle of using AI to support network automation efforts. Those company efforts are based on its Rain platform that was unveiled in 2023, and include services such as Joule and Merlin.

Joule is a closed-loop AI-powered system to create bandwidth-on-demand energy savings at a radio access network (RAN) site. Merlin provides network analytics and business intelligence from the network.

“We did some really unique things to make sure the data we capture and we save is actionable and in real time around a historical basis,” Opanga CEO Cole Brodman previously explained to SDxCentral about its offering. “They can go and draw the information they need out of it to help with capacity planning, future quality-of-service decisions with understanding of what's happening at a real level versus a highly average level on the radio networks from a throughput perspective.”

Broadman at that time had just taken over as head of Opanga, having previously spent more than 11 years at T-Mobile US, where he served as chief development officer, chief technology and innovation officer, and chief marketing officer. He joined Opanga as a board member and advisor in 2015.

Opanga last year added former T-Mobile US CTO Neville Ray to its board.

Opanga also has a standing history with Cisco, with Brodman explaining that the Cisco Ultra Traffic Optimization (CUTO) platform “is their version of our product.”

“It’s the exact same software but built in their mobile packet core,” Brodman said.

That relationship includes Cisco Investments injecting money into Opanga’s operations.

“As we look toward the future of AI-powered network optimization, we’re always on the lookout for innovative solutions that have the potential to transform the telecom industry,” Cisco Investments VP Aleem Rizvon said about that investment. “As part of our initiative to support companies who foster innovation and fund secure and reliable AI solutions, we’re excited to invest in Opanga as it continues to help make networks smarter, faster and more sustainable.”