Fractile

Fractile is building chips, systems and software to radically improve the speed and cost of running frontier AI inference. Existing hardware is good for training frontier AI models, but very poorly suited to subsequent inference of the trained model, which is increasingly defining the new era of AI compute workloads. Fractile’s revolutionary approach to fusing computation with memory eliminates this bottleneck, and can scale to allow running the world’s largest models at a global scale. We've rearchitected the inference stack from the ground up - silicon to software - to enable generational leaps in capability on the world’s hardest and most important problems.