Rebellions

rebellions.ai

Rebellions builds and delivers AI accelerator chips and data center systems designed for efficient inference in large-scale applications. It offers products such as the ATOM™ family, which includes Lite, Max, and rack-scale configurations along with next-generation chiplet-based accelerators like REBEL-Quad featuring high-bandwidth HBM3E memory. Rebellions also provides a full-stack software ecosystem, including an SDK and compiler that support PyTorch 2.0 and vLLM for seamless deployment of models. Its architecture emphasizes performance-per-watt and energy efficiency while scaling across data center deployments. The company operates in the AI chip and inference infrastructure space and serves customers deploying generative AI workloads.