Akamai has rolled out Nvidia's AI Grid reference design across its network of 4,400 Edge locations.

The Nvidia AI Grid sees Nvidia hardware deployed at the network Edge, and has already been adopted by several telcos.

According to Akamai, the company is the first to operationalize the offering and is rolling out "thousands" of the Nvidia RTX Pro 6000 Blackwell Server Edition GPUs across its global network of locations.

The company says this will enable enterprises to run agentic and physical AI "with the responsiveness of local compute and the scale of the global web."

"AI factories have been purpose-built for training and frontier model workloads — and centralized infrastructure will continue to deliver the best tokenomics for those use cases," said Adam Karon, COO and general manager of the Cloud Technology Group, Akamai. "But real-time video, physical AI, and highly concurrent personalized experiences demand inference at the point of contact, not a round trip to a centralized cluster.

"Our AI Grid intelligent orchestration gives AI factories a way to scale inference outward — leveraging the same distributed architecture that revolutionized content delivery to route AI workloads across 4,400 locations, at the right cost, at the right time."

The 'Edge' has long been posited as having an important role to play in AI inferencing workloads, and seems to finally be capturing the industry's interest.

Akamai claims that the distributed nature of the solution will enable customers to "access fine-tuned or sparsified models" across its global Edge footprint, which offers "massive cost and performance advantage for the long tail of AI workloads."

Suggested use cases include gaming studios seeking to deliver AI-driven NPC interactions, financial institutions offering fraud detection and marketing recommendations rapidly, and broadcasters being able to transcode and dub content in real time.

“New AI-native applications demand predictable latency and better cost efficiency at planetary scale,” said Chris Penrose, global VP of business development - telco at Nvidia. “By operationalizing the Nvidia AI Grid, Akamai is building the connective tissue for generative, agentic, and physical AI, moving intelligence directly to the data to unlock the next wave of real-time applications.”

Telcos adopt Nvidia AI Grid

Akamai is not alone in its enthusiasm for the Nvidia AI Grid offering. A slew of telcos have announced similar partnerships to come.

AT&T is teaming up with Cisco for the "Cisco AI Grid with Nvidia," which will combine AT&T's dedicated IoT core with Cisco's Mobility Services Platform, and deploy the Nvidia AI Grid offering based on the RTX Pro 6000 Blackwell Server Edition GPUs.

“Cisco is proud to work with AT&T to spearhead the shift toward physical AI,” said Masum Mir, SVP and GM, provider mobility, Cisco. “Enabled through Cisco’s Mobility Services Platform and extended with Cisco’s AI Grid with Nvidia, AT&T can bring real-time AI inference closer to where data is generated—unlocking new opportunities while enabling enterprises, AI inferencing Independent Software Vendors, and developers to innovate at the network Edge.”

Spectrum, part of Charter Communications, is also set to offer the Nvidia AI Grid, deploying the servers at Edge locations along its broadband network. The company notes that its network has proximity to studios, making it ideal for animation artists looking to render blockbuster-level CGI with GPU compute resources.

Spectrum says it will be able to scale up to hundreds of megawatts of power across 1,000 Edge data centers and hubs with latencies of less than 10 milliseconds - and in some cases, less than five milliseconds.

Comcast, too, is bringing Nvidia AI Grid to its "regional facilities" and has thus far launched a "field trial." DCD's sister publication SDx has done a dive into the project.

In an alternative announcement, tower company Crown Castle is working with Available Infrastructure on its own Edge AI network.

As reported by Network World, the two are partnering on a $5bn initiative dubbed "Project Qestral" that will see the establishment of 1,000 AI Edge locations in 100 US cities by the end of the year. The deployments will be colocated at Crown Castle's telecom sites.

“Our strategy is to industrialize and modularize deployment by building on telecom co-location and pre-existing physical infrastructure rather than greenfield hyperscale construction,” said Dan Medina, EVP of strategy at Available Infrastructure.

Crown Castle owns, operates, and leases more than 40,000 cell towers and roughly 90,000 miles (144,841km) of fiber.