Rebellions secured $400 million in a pre-IPO funding round led by Mirae Asset Financial Group and the Korea National Growth Fund, bringing total funding to $850 million and valuing the company at approximately $2.34 billion. The round follows a $250 million Series C in September 2025 and supports the company’s expansion into the U.S. market, scaling of its Rebel100 platform, and preparation for a future IPO. The investment underscores growing demand for efficient AI inference infrastructure as deployment economics and power constraints become central to AI adoption.
The company introduced RebelRack and RebelPOD, modular infrastructure systems designed for production-scale AI inference deployments. RebelRack provides a fully integrated rack-level compute unit, while RebelPOD aggregates multiple racks into scalable clusters for large-scale environments. Both systems are built on Rebellions’ chiplet-based Rebel100 NPU and optimized for performance-per-watt and cost efficiency in real-world data center conditions. The company positions these offerings as turnkey infrastructure that can be deployed and replicated across cloud providers, neoclouds, telecom operators, and government environments.
Rebellions’ strategy centers on a software-defined, open ecosystem approach to AI infrastructure. Its cloud-native stack integrates with Kubernetes and widely adopted frameworks such as vLLM, PyTorch, Triton, and Hugging Face, enabling distributed inference and heterogeneous deployment without proprietary lock-in. The company has expanded its global leadership team with U.S.-focused growth initiatives led by Marshall Choy, targeting demand for efficient inference platforms that extend existing data center investments while supporting new AI workloads.
- $400M pre-IPO round led by Mirae Asset Financial Group and Korea National Growth Fund
- Total funding reaches $850M; valuation estimated at $2.34B
- Launch of RebelRack (rack-scale) and RebelPOD (cluster-scale) AI inference systems
- Built on Rebel100 NPU with chiplet architecture optimized for performance-per-watt
- Software stack integrates Kubernetes, vLLM, PyTorch, Triton, Hugging Face, and OpenShift
- Focus on U.S. expansion across cloud, neocloud, telecom, and government sectors
- Korea National Growth Fund selects Rebellions as first investment under K-Nvidia initiative
“AI is now measured by its ability to operate in the real world – at scale, under power constraints, and with clear economic return,” said Sunghyun Park, Co-Founder and CEO of Rebellions.
🌐 Analysis: Rebellions aligns with a broader shift toward inference-optimized infrastructure as hyperscalers and enterprises prioritize cost efficiency and power constraints over raw training performance. Competitors including NVIDIA, AMD, and a growing cohort of AI ASIC startups continue to expand vertically integrated offerings, while modular rack- and pod-level systems are emerging as a key deployment model for scaling AI across heterogeneous data center environments.
