Building the next generation of AI inference hardware, Fractile has closed a $220.0M Series b funding round. The company is developing chips, systems, and software designed to radically improve the speed and cost of running frontier AI models.
Fractile aims to solve a critical bottleneck in the AI compute landscape: existing hardware, while effective for training, is poorly suited for the subsequent inference of trained models. By rearchitecting the inference stack from silicon to software, Fractile’s approach fuses computation with memory, enabling generational leaps in capability for the world’s largest AI models.
