Ukraine Strikes Expose Russia’s AI Hardware Dependence

Three Yandex supercomputers used to train artificial intelligence models were reportedly damaged in Ukrainian attacks on Russian data centres earlier this month. The damage remains unclear, and Yandex hasn’t commented. A war between states now also hits computing infrastructure that governments and companies depend on to compete in AI.

Independent Russian outlet T-Invariant quoted experts in supercomputer infrastructure who said all three systems suffered serious damage and would remain out of service for an indefinite period. Yandex hasn’t confirmed their account.

The infrastructure under fire

The Yandex data centre in Sasovo houses the Chervonenkis and Lyapunov supercomputers. It came under attack on October 8. Satellite images showed a major fire and a burned section of the building, T-Invariant reported. Yandex’s third supercomputer, Galushkin, sits in Vladimir. A Ukrainian attack on October 11 forced the data centre there to cease operations entirely.

The centres use free cooling, which pulls outside air through ventilation systems. So, fire, smoke and combustion byproducts could damage equipment beyond the immediate fire zone, according to the article. The machines’ precise condition remains unknown.

Ukraine also struck a Sber data centre in Balakovo, Saratov Oblast, on October 14. The article lists two other Russian supercomputers, Christofari Neo and Christofari. SberCloud built them, and they remain at Sberbank’s data centre in Skolkovo, outside Moscow. That centre hasn’t been attacked.

These machines aren’t just entries on a technology scoreboard. They process enormous datasets and now play an important role in training large neural-network models. Yandex said it designed its systems specifically for machine-learning tasks and used them to train YandexGPT. Christofari and Christofari Neo helped develop Russian-language and generative models, including the GigaChat family and earlier projects.

A race built on imported parts

In November 2021, seven Russian supercomputers appeared on the TOP500 list of the world’s most powerful computing systems. Chervonenkis ranked 19th, Galushkin 36th and Lyapunov 40th. Christofari Neo ranked 43rd, while Christofari placed 72nd.

Russia built infrastructure, software and applications of its own, but relied exclusively on imported hardware components for computing power. All five Russian systems in the TOP500 use American AMD or Intel processors and NVIDIA A100 or V100 accelerators. After Russia’s full-scale invasion of Ukraine in February 2022, many Western countries imposed export restrictions that limited access to semiconductors, equipment and other high-tech components. Even legal supplies of older components became significantly harder to obtain.

The systems have fallen behind as other countries and major technology companies adopted newer accelerators. Chervonenkis still delivers about 21.53 petaflops, but its ranking dropped from 19th in 2021 to 101st in the June 2026 rating. Russia now has five systems in the TOP500 and ranks 21st by number of systems. The June 2026 leader, LineShine, reached 2.198 exaflops—more than 100 times Chervonenkis’s performance.

The price of state competition

To repair a damaged system, operators need compatible graphics processors, processors, networking equipment and memory. A modern replacement takes access to the next generation of accelerators. Russia can run large computing complexes and build competitive models, but it lacks an industrial counterpart to the NVIDIA A100 and access to more advanced H100 and H200 accelerators.

The wider AI race centres on the United States and China, whose supercomputing infrastructure is larger and updated rapidly. The Stanford AI Index 2026 said American companies created 50 of the most notable new AI models in 2025, while Chinese companies created 30. That year, private AI investment reached approximately $286 billion in the United States and $12.4 billion in China; those figures don’t include a significant portion of Chinese government funding.

Russia ranked 31st among 83 countries in the Global AI Index 2024. At the time, a Sberbank representative estimated that Russian generative models lagged those of the United States and China by approximately six to nine months. Significant damage to Russian supercomputers could make improving that position harder. The machines, components and rankings tell a blunt story: governments and corporations compete for technological power, while war and export restrictions shape who can build, repair and upgrade the systems.