Address

20 Rochdelskaya St., Moscow

Artificial Intelligence Infrastructure

This report examines the key characteristics of AI infrastructure and major global trends, with a focus on the specific conditions of the Russian market. It focuses on three interrelated components of AI infrastructure: energy, computing, and connectivity. To assess these components holistically, the report develops a total cost of ownership model for AI infrastructure facilities. Drawing on international experience and an analysis of the Russian market, the report presents a set of policy recommendations to support the continued development of AI infrastructure in Russia.

Key findings


  • AI is driving the emergence of a new class of infrastructure: AI data centers. These facilities have fundamentally different computing and connectivity requirements than traditional IT infrastructure.

  • Power constraints and lengthy grid connection timelines are major barriers to AI data center development worldwide.

  • Specialized AI accelerators are increasingly scarce and becoming obsolete at a rapid pace.

  • The network infrastructure for AI data centers requires fundamentally different solutions both within individual facilities and for connectivity between facilities.

  • Russia faces the same global challenges across all components of AI infrastructure, with the effects further compounded by sanctions.

  • Through 2036, the development of AI infrastructure in Russia could follow one of three scenarios: technological sovereignty, adaptive integration, or infrastructure constraints.

  • A combination of regulatory, cross-cutting, and targeted policy measures, together with strategic infrastructure siting, would create the conditions for Russia to achieve the technological sovereignty scenario.

 

This analysis draws on a review of global and Russian practices, a qualitative comparison of Russia’s competitiveness with that of China, the EU, and the US, a quantitative analysis using a total cost of ownership model for AI data centers, and a spatial assessment of potential sites for AI infrastructure across Russia.

Authors

We use cookies to improve the site