The key facts

  • Ukrainian drones struck three Yandex data centers on October 8, 9, and 11, causing fires and service disruptions in Russia and CIS countries, including Uzbekistan.
  • The attacked centers include one of Russia's largest data processing facilities with a design capacity of 63 megawatts.
  • Three leading supercomputers used for AI training — Chervonenkis, Lyapunov, and Galushkin — may have been completely disabled; the exact extent of the damage is unknown.

Last week, Ukrainian drones attacked several of Yandex's data centers. The strikes may have caused serious damage to the company's databases. As a result, disruptions were observed in some of Yandex's services in Russia and CIS countries, including Uzbekistan.

Attack on three data centers

Initially, on the night of October 8, the company's data center located in the town of Sasovo, Ryazan region, was struck. This facility, opened in 2014, is one of Yandex's main data processing centers. Its power capacity reaches up to 12 megawatts, and its infrastructure contains tens of thousands of servers. After the drone strike, a major fire broke out and the center's operations were temporarily halted. This caused disruptions in the operation of various services that use Yandex Cloud resources.

On October 9, another data center located in the Grabsevo industrial park in the Kaluga region came under attack. The company itself described this facility, commissioned in 2022, as one of the largest data processing centers in Russia. The center's design capacity is 63 megawatts. As a result of the attack, several modules of the facility were completely disabled.

On the night of October 11, Ukrainian drones also attacked Yandex's data center in the city of Vladimir. The successive strikes affected the operation of digital services not only in Russia but also in other CIS countries. In particular, problems arose in Uzbekistan when using some of Yandex's services.

The scale of the damage

According to preliminary analyses, the damage to the data centers may deal a serious blow to the development of artificial intelligence technologies in Russia. One of the biggest losses cited is the failure of three leading supercomputers used for training neural networks, large language models, and machine vision systems.

The Chervonenkis and Lyapunov supercomputers in Sasovo, and the Galushkin supercomputer in the Kaluga region, were used for the company's artificial intelligence developments, in particular for training neural networks. Notably, Chervonenkis is described as the most powerful supercomputer built in Russia in the last 15 years. When it was launched in 2021, it ranked 19th in the global TOP500 list of the most powerful supercomputers. Galushkin was also designed for tasks requiring high computing power, such as training AI models. As a result of the attacks, all three supercomputers may have been completely disabled. However, there is no full information about their condition or the exact extent of the damage.