Rebellions and Red Hat Introduce Red Hat OpenShift AI Powered by Rebellions NPUs to Fuel Choice and Flexibility in Enterprise AI

SEOUL, South Korea — December 11, 2025 — Rebellions and Red Hat, the world’s leading provider of open source solutions, today announced Red Hat OpenShift AI powered by Rebellions neural processing units (NPUs), integrating Red Hat’s scalable, flexible and open source AI inference capabilities with Rebellions’ energy-efficient NPUs to deliver a validated full-stack enterprise AI platform. This collaboration fully supports Red Hat’s goal to deliver “any model, any accelerator, any cloud” to customers globally by enabling more choice and optionality in the architectures underpinning AI workloads.

As organizations move AI initiatives into actual business use cases, challenges such as infrastructure costs, deployment complexity, and the need for flexible environments with enhanced security footprints are becoming more prominent. GPU-centric environments on their own can be insufficient to meet the performance and efficiency demands at scale. Red Hat OpenShift AI powered by Rebellions NPUs addresses these challenges by enabling organizations to operate AI workloads more efficiently across environments.

Rebellions’ NPUs are designed with an architecture optimized for AI inference, delivering enhanced energy efficiency compared to traditional GPUs, more effectively reducing data center deployment and operational costs at both the server and rack levels. Their specialized full-stack software and support for leading open source AI frameworks provide the same level of convenient development environment to that of GPUs.

Red Hat OpenShift AI powered by Rebellions NPUs offers a fully integrated solution, from hardware to model serving, that has been validated by Red Hat and Rebellions for enterprise-grade compatibility. Rebellions’ software stack runs natively on Red Hat OpenShift AI, reducing overhead and accelerating deployment. The Rebellions NPU Operator, certified for Red Hat OpenShift,  enables more seamless integration and trusted support across on-premises and multi-cloud environments, helping to meet data sovereignty and regulatory compliance. This solution provides:

• Enterprise-ready AI at scale: Run large language models (LLMs) and inference workloads with high throughput, low latency and superior power efficiency, leveraging vLLM integrated with Rebellions’ rack-scale NPU solutions for distributed processing.
• Secure and compliant: Keep data on-premises and meet regulatory requirements with Red Hat’s trusted platform and Rebellions’ secure hardware.
• Simplified operations: Manage NPUs like GPUs, using Red Hat’s unified platform, reducing complexity and accelerating adoption.
• Flexible and scalable: Deploy where your data lives, from core to edge, with linear scale-out.

More information on Red Hat OpenShift AI powered by Rebellions NPUs can be found at the Red Hat Ecosystem Catalog.

Supporting Quotes

Brian Stevens, senior vice president and chief technology officer for AI, Red Hat

“The future of enterprise AI demands architectural choice beyond proprietary, monolithic stacks. Our collaboration with Rebellions is a powerful step in delivering Red Hat’s ‘any model, any accelerator, any cloud’ strategy. By tightly integrating the open, scalable capabilities of Red Hat OpenShift AI with Rebellions’ energy-efficient NPUs, we’re giving enterprises a validated, full-stack alternative. This enables customers to deploy their most demanding AI inference workloads with the required efficiency, low latency, and horizontal scalability that is critical for production AI.”

Sunghyun Park, CEO, Rebellions

“As AI serving and inference accelerate, enterprises need practical infrastructure that meets their requirements for performance, cost efficiency, and data sovereignty. Through this collaboration, Red Hat and Rebellions will provide a validated, end-to-end inference platform that replaces the fragmented approaches of the past. This will help enterprises scale their AI services more efficiently and securely, while also presenting a new NPU-based alternative to the traditionally GPU-centric environment.”