
Russia Unveils First Domestic AI Accelerator for Offline Neural Networks
LinQ HX operates locally, runs multiple models simultaneously, and consumes no more than 40W

LinQ HX operates locally, runs multiple models simultaneously, and consumes no more than 40W

Increased supply directly fuels interest

Earlier, the State Duma adopted a law differentiating between "sovereign" and "national" models

The system will check hundreds of thousands of transactions and should identify suspicious schemes faster

For now, it concerns audio and video materials

The service will consider opening hours, menu, parking, and other conditions in one request

Copyright holders fear their works will become free fuel for training models

The marketplace wants to make choices more transparent amid the spread of generative graphics

The model processes long documents up to four times faster and is almost twice as compact as the previous version

Operators are starting to implement neural networks - connection speed could increase by up to 30%

New powers could affect almost 28,000 IT companies as early as September 1st

Changes affected computer science, technology, and "Fundamentals of Safety and Homeland Defense" courses