Rebellions and Standard Energy Partner to Develop AI for Data Center Energy Storage Management December 8, 2025 As the demand for artificial intelligence computing surges, data center operators face mounting pressure to manage energy consumption and costs. A critical challenge lies in the unpredictable power spikes inherent to AI workloads, which can trigger severe financial penalties under standard utility contracts when peak demand exceeds agreed capacity. In a strategic move to address this issue, South Korean AI chip startup Rebellions and Singapore-based battery firm Standard Energy have initiated a joint development project for an artificial intelligence system designed to control energy storage systems (ESS) within AI data centers. The collaboration, first reported by BusinessKorea, aims to create an AI that can manage power flow between servers and ESS in millisecond intervals by monitoring real-time demand and automatically controlling power discharge. The project is an extension of the companies' existing "Dopamine" integrated solution, which combines Rebellions' neural processing unit (NPU) servers with Standard Energy's vanadium-ion battery ESS. The new control AI will be integrated into this platform. The partners claim the technology will allow operators to supply stored battery power during demand peaks, thereby avoiding capacity overages that can incur surcharges ranging from 1.5 to four times the base electricity rate. A functional prototype of the AI control solution is targeted for completion in the first half of 2026, with commercial deployments to follow. Rebellions, a specialist in energy-efficient AI inference accelerators founded in 2020, was recently valued at $1.4 billion following a Series C funding round. Standard Energy, established in 2019, designs and manufactures vanadium-ion battery systems for energy storage. This development signals a growing industry trend towards leveraging AI not just as a workload, but as a critical tool for infrastructure optimization. By intelligently bridging compute and energy storage, such solutions could significantly enhance the economic viability and operational stability of next-generation AI data centers, particularly in regions with high or volatile energy costs. Source: datacenterdynamics
Rebellions and Standard Energy Partner to Develop AI for Data Center Energy Storage Management
