Akamai Technologies has officially integrated Nvidia AI Grid\ntechnology into its infrastructure. This technology enables the\ndistribution of artificial intelligence (AI) computing to thousands of\nedge points worldwide.<\/p>\n
This step positions Akamai as the first company to implement Nvidia\u2019s\nAI Grid reference design on a global scale. The integration is part of\nthe development of the Akamai Inference Cloud platform, which was\npreviously introduced.<\/p>\n
Through this approach, Akamai deploys specialised AI workloads,\nparticularly inference, to more than 4,400 edge locations, as well as\nregional data centres and the core. The goal is to balance latency,\ncost, and performance in AI processing.<\/p>\n
Akamai\u2019s Chief Operating Officer, Adam Karon, stated that the\npreviously centralised AI infrastructure model is no longer sufficient\nto meet the needs of modern applications.<\/p>\n
\u201cOur intelligent AI Grid orchestration provides a way for AI\nfactories to extend inference outwards, leveraging the same distributed\narchitecture that has revolutionised content delivery to direct AI\nworkloads to 4,400 locations, at the right cost and at the right time,\u201d\nsaid Karon.<\/p>\n
The GPUs complement Akamai\u2019s edge network, allowing the company to\noffer a combination of high-powered centralised computing and\nuser-proximate processing.<\/p>\n
At the edge, the system handles requests requiring low latency.\nMeanwhile, heavy workloads such as model training and advanced\nprocessing remain in the core data centres.<\/p>\n
Akamai emphasises that the core of the AI Grid is an intelligent\norchestrator that manages workload distribution in real-time. The system\noptimises what is termed \u201ctokenomics\u201d, namely cost efficiency per token,\nresponse time, and throughput.<\/p>\n
With mechanisms like semantic caching and intelligent routing, AI\nrequests are directed to the most suitable resources. Lightweight\nworkloads can run at the edge, while heavy tasks are allocated to\nhigh-end GPUs.<\/p>\n
This approach is claimed to reduce operational costs while improving\nAI service performance, especially for large-scale applications.<\/p>\n
Akamai notes that initial adoption of the platform has been high\nacross various sectors requiring rapid responses.<\/p>\n
In the gaming industry, the technology is used to deliver real-time\nAI-based non-player character (NPC) interactions with latency under 50\nmilliseconds.<\/p>\n
In the financial sector, the network is utilised for fraud detection\nand service personalisation as users enter applications.<\/p>\n
Meanwhile, in the media industry, broadcasters use the technology for\ndirect content transcoding and dubbing processes.<\/p>\n
Usage is also expanding into the retail sector, particularly for\nin-store AI applications and AI-based checkout systems.<\/p>\n
Distributing computing to the edge is seen as a solution to overcome\nlatency and scalability challenges, while enhancing user experience.<\/p>\n
The Akamai Inference Cloud platform is currently available to select\ncorporate customers. The company has also announced securing service\ncontracts worth $200 million over four years for providing GPU clusters\nin edge AI infrastructure.<\/p>", "url": "https:\/\/jawawa.id\/newsitem\/akamai-integrates-nvidia-ai-grid-distributing-ai-computing-to-the-edge-1773899478", "image": "" }, "sponsor": "Okusi Associates", "sponsor_url": "https:\/\/okusiassociates.com" }