Red Hat and Rebellions Unveil New AI Platform

Red Hat collaborates with Rebellions to introduce a full-stack enterprise AI inference platform with zero emissions of power.

Red Hat and Rebellions have introduced Red Hat OpenShift AI together, which is a full-stack enterprise platform on next-generation neural processing units offered by Rebellions. The relationship will improve the predilection of Red Hat to provide flexible AI infrastructure by providing more architectural options in models, accelerators, and cloud environments to the enterprise. This launch also puts Red Hat in a position to move forward in its approach to offering global customers scalable, secure and efficient AI deployment.

Over the past years, as large businesses start to deploy real-world AI applications, weaknesses in the infrastructure of GPU-based systems have become more constraining, such as rising infrastructure expenses, increasing implementation complexity, and security necessities. The new platform also takes advantage of Rebellion's NPUs by providing better energy efficiency and lowering operation overheads across data centers. The stack was proven to be free of bugs and works on Red Hat OpenShift AI, which guarantees easier deployment, integration, and adherence to the requirements of the data sovereignty.

The solution provides high throughput and low latency inference for large language models without compromising enterprise-level security. Red Hat allows organizations to manage NPUs like GPUs using their unified framework and has made the process of adoption easier and faster.

Red Hat leaders have underlined that the collaboration will be a huge step to follow through with the open and modular AI infrastructure. Rebellions also noted that the unified platform offers an effective alternative to the traditional GPU environments, allowing the scaling of AI services to be done efficiently and securely.

The partnership makes Red Hat OpenShift AI a competitive offering to the organizations wanting to have a flexible and cost-effective AI infrastructure in the core, cloud, and edge environments.