Akamai Technologies, Inc. (NASDAQ: AKAM) has announced Akamai Inference Cloud, a groundbreaking platform transforming how and where artificial intelligence is applied by extending inference processes from traditional data centers to the Internet Edge.
This solution enables the execution of advanced and autonomous AI inferences directly at the Edge, bringing processing closer to users and their devices. Unlike traditional architectures, Akamai Inference Cloud is designed to deliver real-time AI processing with low latency and global reach. The launch combines Akamai’s extensive experience in distributed architectures with NVIDIA Blackwell’s AI infrastructure technology, redefining accelerated computing to unlock AI’s full potential.
The rising demand for next-generation smart applications—such as personalized digital experiences, autonomous assistants, or real-time decision systems—requires inference to be performed as close as possible to the user. This guarantees instant responses and efficient request handling based on local context.
Designed to meet these needs, Akamai Inference Cloud functions as a generative, distributed Edge platform, bringing the power of NVIDIA AI to where data is generated and decisions are made, providing an infrastructure prepared for the future of global AI.
“The next wave of AI demands the same user-centric approach that enabled the Internet to become the ubiquitous global platform it is today,” states Dr. Tom Leighton, CEO and co-founder of Akamai. “Akamai has already addressed this challenge and will do so again. Thanks to NVIDIA’s AI infrastructure, Akamai Inference Cloud will meet the growing demand to scale AI inference capacity and performance, bringing decision-making to thousands of locations worldwide, enabling faster, smarter, and more secure responses.”
“Inference has become the most computationally intensive phase of AI, requiring real-time reasoning at a planetary scale,” says Jensen Huang, founder and CEO of NVIDIA. “Together, NVIDIA and Akamai are bringing inference closer to users around the world, enabling faster, more scalable generative AI, and opening the door to the next generation of smart applications.”
This solution redefines where and how AI is used by bringing intelligent, autonomous inference directly to users and their devices. The platform combines NVIDIA RTX PRO servers, equipped with NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs, NVIDIA Bluefield-3 DPU, and NVIDIA AI Enterprise software, with Akamai’s distributed cloud infrastructure and global Edge network, spanning over 4,200 locations worldwide. Akamai Inference Cloud will drive Akamai’s vision of highly scalable, distributed AI performance worldwide by leveraging the latest NVIDIA technologies like the recently announced Bluefield-4 DPU, further accelerating and securing access to data and inference workloads from core to Edge.
This partnership pushes the boundaries of AI with Akamai Inference Cloud:
- Extending enterprise AI factories to the Edge to enable intelligent commerce agents and personalized digital experiences: AI factories are powerful tools coordinating the entire AI lifecycle, from data ingestion to large-scale intelligence creation. Akamai Inference Cloud expands AI factories to the Edge by decentralizing data and processing, routing requests to the best models via Akamai’s extensively distributed Edge locations. This allows intelligent agents to adapt instantly to user location, behavior, and intent, acting autonomously to negotiate, formalize, and optimize transactions in real-time.
- Enabling agents and streaming inference to provide instant financial insights and make real-time decisions: AI agents require multiple sequential inferences to complete complex tasks, which can introduce delays hindering user interaction. Autonomous AI workflows involve multiple inference calls, and network delays in each call can make responses appear slow or too late for machine-to-machine latency requirements. Akamai Inference Cloud’s native Edge architecture offers near-instant responses, allowing AI agents to operate with human-like responsiveness in multi-step workflows. This can be invaluable for fraud detection, accelerating secure payments, and enabling high-speed decision-making in industrial sectors.
- Allowing real-time physical AI systems to operate beyond human response capabilities: Physical AI systems, such as autonomous vehicles, industrial robots, and smart city infrastructure, require millisecond-level decision-making to safely interact with the physical world. Akamai Inference Cloud is designed to enable physical AI to process sensor data, make safety decisions, and coordinate actions at speed, transforming everything from manufacturing plants and delivery drones to surgical robots and autonomous traffic networks into intelligent, responsive systems that can operate safely alongside humans.
- Accelerating the return on investment: Managing complex, distributed AI workloads across multiple cloud regions requires specialized knowledge and teams. Akamai Inference Cloud’s intelligent orchestration layer automatically directs AI tasks to optimal locations. Routine inference runs instantaneously at the Edge via NVIDIA’s NIM microservices, while sophisticated reasoning leverages centralized AI factories, all managed through a unified platform that simplifies infrastructure complexity.
Akamai Inference Cloud is initially available in 20 locations worldwide, with plans underway to expand deployment.