Urgent.News

What's breaking now, across thousands of outlets.

AI

Internal Amazon documents reveal a major overhaul of one of the world's biggest AI data centers

Amazon is redesigning part of its Indiana AI campus into a massive Trainium-powered cluster for its next frontier AI model, internal documents show.

Abstract editorial illustration

Amazon is undertaking a major overhaul of its Indiana data center campus to support its Artificial General Intelligence (AGI) team's ambitious frontiers model project. This development comes despite recent job cuts within Amazon's AGI organization and the winding down of the previous Nova model series. The company is increasing its AI infrastructure investments, with a target of $220 billion for AI capital expenditures by 2026, in response to the rising demand for AI computing power and escalating component prices.

Internal planning documents, reviewed by Business Insider, reveal an effort to consolidate multiple data centers and deploy thousands of Trainium-powered AI servers. The initiative, dubbed "AGI Pivot," aims to strengthen Amazon's AGI organization and its future in-house AI models. According to people familiar with the matter, the company is striving to deploy a substantial number of Trainium-powered AI servers, with an "emergent request" to deploy over 6,000 of these servers.

The redesign of data centers is part of a broader initiative to maximize returns on Amazon's AI investments. By integrating existing data centers into a more efficient AI cluster, Amazon can optimize the use of its existing infrastructure, which is expected to generate long-term attractive returns. This approach aligns with a statement by CEO Andy Jassy, who emphasized that AI infrastructure should yield attractive long-term returns, given the long-lasting productivity of data centers compared to the refreshable nature of servers and networking equipment.

The Indiana campus, which hosts facilities for Project Rainier, a Trainium-powered AI supercomputer primarily designed for Anthropic, will not be affected by the AGI initiative. However, the project will create a larger, more efficient AI system, known as the AGI SuperCluster, by redesigning networking, storage, and fiber-optic infrastructure. This will allow multiple data centers to function as a unified AI system, thereby maximizing performance.

To achieve this, Amazon plans to replace older Trainium 2 systems with newer Trainium 3 servers. Additionally, the company is converting some existing buildings into "annexes" that can share core networking equipment with neighboring data centers, thus functioning as one interconnected AI system. Despite the recent layoffs within the AGI organization, the push for frontier AI remains resilient, with Amazon significantly expanding its AI investment and raising its projected 2026 capital expenditures to $220 billion.

Written by urgent.news from Business Insider's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Also reported by 3 other outlets

Read the original at businessinsider.com →

More in AI

More from Thursday 6 August →