Akamai & NVIDIA launch global edge AI platform for real-time use

Akamai and NVIDIA have announced a partnership to launch Akamai Inference Cloud, a distributed platform designed to process AI tasks at the edge of the internet using the NVIDIA AI technology stack.

The Akamai Inference Cloud aims to decentralise AI inference, which is the process of applying trained machine learning models to new data in real-world scenarios, to provide real-time predictions and automated decision-making. By moving AI from centralised data centres closer to where data is generated, the platform is built to deliver low-latency, scalable AI performance worldwide.

Deployment at the edge

The partnership enables AI inference to be performed at the edge of the network, reducing response times and allowing organisations to process and analyse data near its point of creation. According to Akamai, this is increasingly necessary for modern use cases such as digital experiences, smart agents, and applications requiring rapid, context-aware decisions.

Key features highlighted by both companies include a decentralised approach that routes AI tasks to optimal models and locations, as well as capabilities for rapid, human-like responses in complex workflows. The architecture is designed to support real-time processing in industrial settings, such as autonomous vehicles or robotics, where milliseconds matter for safety or performance.

Akamai Inference Cloud is initially available in 20 locations across the globe, with expansion plans underway. The service leverages Akamai's global edge network, which includes more than 4,200 sites, combined with NVIDIA's AI hardware and software portfolio.

Combining infrastructure and expertise

The platform integrates NVIDIA RTX PRO Servers, including RTX PRO 6000 Blackwell Server Edition GPUs and NVIDIA BlueField-3 DPUs, with Akamai's globally distributed computing resources. NVIDIA AI Enterprise software provides the AI capabilities within this infrastructure. Akamai stated that it will also adopt the newly announced NVIDIA BlueField-4 DPU to further enhance and secure data flows and AI workloads at the edge.

"The next wave of AI requires the same proximity to users that allowed the internet to scale to become the pervasive global platform that it is today. Akamai solved this challenge before - and we're doing it again. Powered by NVIDIA AI infrastructure, Akamai Inference Cloud will meet the intensifying demand to scale AI inference capacity and performance by putting AI's decision-making in thousands of locations around the world, enabling faster, smarter, and more secure responses," said Dr. Tom Leighton, Akamai CEO and co-founder. 

The companies emphasise that the platform will allow enterprise AI systems-such as personalised recommendation engines and intelligent customer service agents-to operate near users, tailoring responses in real time based on location and need. This is intended to improve customer engagement, support autonomous decision-making, and reduce the time required for complex data processing workflows.

"Inference has become the most compute-intensive phase of AI - demanding real-time reasoning at planetary scale. Together, NVIDIA and Akamai are moving inference closer to users everywhere, delivering faster, more scalable generative AI and unlocking the next generation of intelligent applications," said Jensen Huang, founder and CEO, NVIDIA. 

Use cases and benefits

Akamai Inference Cloud is set to support a range of emerging AI-driven applications. These include extending AI "factories" to the edge for use in commerce and personalisation, enabling smart agents to act based on local context and user intent, and supporting autonomous negotiation or transaction optimisation activities in real time.

For financial and industrial clients, the companies highlight the benefits of streaming inference and agent-based workflows enabled by edge computing. This approach allows AI systems to deliver instant decision-making, which can be used in scenarios such as fraud detection, secure payments, and high-speed industrial automation.

In the field of physical AI, the platform is intended to power systems like industrial robots, smart city infrastructure, and autonomous vehicles, which require rapid sensor data analysis and split-second safety decisions. Akamai claims its platform can coordinate actions for these systems safely and responsively at the speed required for real-world operation.

Intelligent orchestration

The orchestration layer in Akamai Inference Cloud is designed to automate task management for AI workloads. Routine inferences can be handled instantly at the edge, while more complex reasoning can be processed in centralised AI factories. The platform provides unified management, which Akamai states will simplify infrastructure requirements for clients.

The initial deployment of Akamai Inference Cloud covers 20 sites, with additional expansion planned as demand and use cases grow worldwide. Both Akamai and NVIDIA state that this edge-native approach is necessary to meet increased demand for real-time, distributed AI across a broad spectrum of industries and applications.