NVIDIA, Ineffable Intelligence Team Up to Build the Future of Reinforcement Learning Infrastructure

NVIDIA and Ineffable Intelligence, the AI lab founded by AlphaGo architect David Silver, have announced an engineering-level collaboration to build reinforcement learning infrastructure at scale. Unlike pretraining on fixed human datasets, RL workloads generate data on the fly through continuous act-observe-score-update loops, placing unique demands on interconnect, memory bandwidth, and serving. The joint work begins on NVIDIA Grace Blackwell hardware and will extend to the upcoming Vera Rubin platform, with the goal of enabling agents to discover new knowledge through simulation and experience rather than learning from existing human data.