Acrab raised $350 million to build infrastructure for agentic AI: betting on decentralized computing
Singapore-based startup Acrab has announced raising over $350 million in total funding to develop computing infrastructure for agentic artificial intelligence systems. This is one of the largest rounds in the AI hardware segment in the Asia-Pacific region, highlighting the growing demand for specialized solutions for autonomous AI agents.
Founded in 2024, Acrab focuses on creating a comprehensive full-cycle computing architecture. The development portfolio includes AI chips, systems for local deployment of large language models (LLMs), operating systems, multimodal interfaces, and AI agent management technologies. The company's key product is the GΞLIX platform, designed for local execution of agentic AI tasks without the need for constant cloud server access.
According to the Acrab team, GΞLIX has already been tested in real-world conditions and is at the stage of transitioning to its first industrial deployment and mass production. The funds raised will be directed toward further platform development, research in computing systems, expanding the partner network, and international expansion.
Analytical commentary: From a market perspective, a young startup raising $350 million signals that investors see enormous potential in the segment of decentralized and local computing power for AI. Agentic AI, capable of autonomously making decisions and executing complex chains of actions, requires fundamentally different infrastructure than traditional cloud models. If Acrab succeeds in scaling GΞLIX to industrial volumes, it could significantly reshape the landscape of the AI hardware market, especially in the context of growing demands for data privacy and latency reduction.