AMD deepens AI inference bet with Taalas deal as chip race heats up

Toronto-based Taalas develops specialized silicon designed to reduce computing and memory bottlenecks in AI ‌inference, the ⁠process ⁠of running trained AI models to generate responses or predictions.