
The strike sequence targets the specific constraint that matters most to Russian digital sovereignty: not internet pipes, but the compute layer that trains models and runs payments. Yandex's six data centers were already the only domestic alternative to AWS and Azure for Russian enterprises; three offline forces the rest of the stack onto competitors' infrastructure during the window when Moscow is trying to decouple from Western cloud.
The supercomputer loss is the deeper hit—Sasovo and Vladimir housed the only large-scale training capacity outside of state military labs, which means any new Russian AI model development now queues behind whatever the Kremlin prioritizes for defense. The civilian AI economy just went on rationing.
Yandex Cloud serves Russia's entire e-commerce and financial-services stack; three data centers offline means Moscow now has no redundancy across its primary cloud provider, forcing 64,000 businesses into emergency failover mode while competing providers absorb the load.
The two destroyed supercomputers housed Russia's core AI training infrastructure—losing them doesn't just disrupt services, it halts the only domestic pipeline for large-scale model development that isn't reliant on smuggled chips. Watch whether Yandex's partnerships with Selectel, K2Cloud, and VK Cloud actually absorb the customer migration, or whether the shift becomes permanent—if enterprises defect, Yandex's market position fragments and Moscow loses control of a critical digital chokepoint it had consolidated over the past decade.
How many of Yandex's 64,000 cloud customers actually migrated to competitors, and did any choose non-Russian providers? Did the supercomputer damage extend to stored model weights, or just the training hardware?
Strategic intelligence, synthesized daily — with a public track record. Every call graded against what actually happened.