Akamai Inference Cloud is a distributed AI infrastructure platform designed to run inference closer to users and data.
Points à retenir :
AI infrastructure has reached a critical inflection point where the physical distance between centralized AI factories and end users has become a fundamental bottleneck for real-time applications. In this IDC Link, IDC’s Dave McCarthy explores how Akamai is addressing this problem by overlaying an intelligent orchestration plane onto its geographically distributed cloud. En savoir plus
- Learn how Akamai Inference Cloud is transitioning enterprise AI from pilot projects into production-grade infrastructure.
- Understand how to overcome the core physical and economic bottlenecks of the traditional hyperscale model, such as 100 ms+ round-trip delays and high egress costs of centralized clouds.
- Implement high-concurrency use cases across gaming, financial services, and live media that require ultra-low latency for peak performance.
Foire aux questions (FAQ)
Akamai is deploying thousands of NVIDIA RTX PRO™ 6000 Blackwell Server Edition GPUs, interconnected with NVIDIA BlueField-3® Data Processing Units (DPUs), across more than 4,400 edge locations. An intelligent orchestration layer acts as a rapid broker, routing AI requests to optimal compute tiers based on proximity, cost, and GPU availability.
By deploying AI inference at the network edge, we significantly shorten the physical distance between compute resources and end users. This enables deterministic, ultra-low latency — an essential requirement for real-time applications in sectors like gaming, financial services, and live media.