FuriosaAI Raises $125 Million Series C to Scale AI Computing

FuriosaAI has completed a $125 million Series C funding round to meet global demand for its flagship AI chip RNGD (Renegade) and accelerate its roadmap. With this round, FuriosaAI's total funding raised to date reaches $246 million. The round included participation from KDB (Korea Development Bank), IBK (Industrial Bank of Korea), Keystone Partners, PI Partners, and Kakao Investment, among others.

FuriosaAI's core technology, the Tensor Contraction Processor (TCP) architecture, is optimized for tensor contraction, the fundamental language of deep learning computation. This allows the compiler to generate efficient data flow paths that minimize unnecessary data movement, resulting in lower total cost of ownership (TCO) and flexible deployment environments.

Recently, LG AI Research adopted RNGD to power its EXAONE foundation models. In testing, RNGD demonstrated its capabilities as enterprise-grade hardware, achieving 2.25x higher performance-per-watt in LLM inference compared to existing GPUs.

The funds raised will be invested in the following two key areas:

• Scaling RNGD production and accelerating market entry: strengthening manufacturing capacity to support global customers and securing additional customers
• Next-generation chip development: designing high-performance, ultra-efficient AI computing solutions optimized for agentic AI and reasoning models