Инфраструктура ИИ: облака, ЦОД, альтернативы Nvidia

On October 7, 2026, the business publication CNews hosts a meetup "Infrastructure for AI: Own Data Center, Cloud Services, and Alternatives to Nvidia". The event brings together procurement companies, cloud service providers, IT integrators, and data center operators.

Key Discussion Topics

At the meetup, three approaches to AI infrastructure will be examined:

• Building proprietary data centers — capital investments, equipment requirements, payback periods
• Cloud providers — rental of ready-made computing resources with flexible scalability
• Alternatives to Nvidia graphics cards — sourcing Chinese-made processors (Ascend, Huawei), domestic developments, joint investments in regional data centers

Why This Question Is Relevant in Russia

The country faces an acute shortage of Nvidia equipment due to export restrictions and sanctions. Companies investing in their own AI systems are forced to seek independent solutions. Choosing a path — a proprietary data center, cloud, or hybrid approach — requires understanding economics, technical capabilities, and strategic vision for 3–5 years.

The meetup provides an opportunity to share experiences between companies that have already faced this problem and those just beginning to plan their AI infrastructure.

What This Means

The shortage of Western equipment is changing the game: Russian business is accelerating the transition to independent infrastructure, choosing between Chinese alternatives, cloud services, and consortium investments in regional data centers. Infrastructure discussion is moving from technical backrooms to the business agenda and is becoming critical for any company with AI ambitions.