Axelera AI Secures Over $250 Million to Scale Edge-AI Chip Production

Axelera AI Secures Over $250 Million to Scale Edge-AI Chip Production

Axelera AI, an Eindhoven-based semiconductor company, has secured more than $250 million in fresh funding to expand its global commercial footprint and scale its edge-AI chip production. The latest investment round was led by Innovation Industries, with participation from SiteGround Capital and various funds and accounts, including BlackRock.

This capital injection follows support from existing backers, including Samsung Catalyst Fund, Bitfury, the European Innovation Council Fund, Invest-NL, CDP Venture Capital, SFPIM, and Verve Investments. Since its launch in July 2021, the company has raised over $450 million across equity, grants, and venture debt. Axelera AI intends to use these funds to grow its manufacturing operations, bolster its customer support and partner networks, and further develop its software tools and SDKs.

An “Edge-First” Architectural Strategy

Axelera AI was founded to address the power consumption, heat generation, and cooling limitations inherent in traditional data center-reliant AI models. By focusing on an “edge-first” architecture, the company designs chips intended to run AI tasks directly at the point of data generation.

“Data centres are hitting power and cooling limits, and as analytics move closer to where data is being created, edge AI solutions must operate within strict energy and bandwidth constraints,” said Fabrizio Del Maffeo, CEO and co-founder of Axelera AI. “We designed our architecture from the ground up to overcome these obstacles.”

The company’s hardware stack is built upon the RISC-V instruction set architecture (ISA) and in-memory computing. By utilizing RISC-V, an open, no-royalties-attached standard, Axelera avoids the licensing models associated with traditional chipmakers. Furthermore, its in-memory computing approach performs calculations within the system’s RAM, a design choice intended to minimize the latency typically introduced by storage devices.

Hardware Innovation: The Metis Platform

Central to the company’s product line is the Metis AI platform. The hardware is designed for deployment in sectors such as smart city infrastructure, industrial inspection, retail analytics, and security applications.

To expand its reach, the company recently launched the Axelera AI Mini PC. This platform integrates the Metis M.2 MAX accelerator with an Intel Core Ultra 125H processor. By offloading AI workloads to the Metis accelerator while the CPU handles pre-processing, post-processing, and system management, the unit eliminates the need for a discrete GPU. This configuration reduces power consumption and system complexity, allowing the device to operate within a typical power draw of 3.5W to 11W.

According to Axelera, the system is capable of:

• Processing more than 25 simultaneous 1080p video streams at 20 frames per second.
• Delivering up to 15 TOPS/W (Tera Operations Per Second per Watt).
• Operating in temperatures ranging from 0°C to 40°C without requiring dedicated air-conditioned enclosures.

Software Ecosystem and Market Positioning

Recognizing that hardware alone is insufficient for widespread adoption, Axelera has invested in the Voyager SDK. This software suite includes tools for model optimization, quantization, and runtime performance monitoring. It also features a “Model Zoo” containing over 100 pre-trained models for tasks such as object detection, pose estimation, and license plate recognition. The company is currently manufacturing through partnerships with TSMC and Samsung. Looking ahead, CEO Fabrizio Del Maffeo noted that the company is developing a new generation of products for large language models and large multimodal models, with plans to unveil this family later this year and move into full production in 2025. “Axelera is solving one of the most fundamental constraints in Edge AI adoption: the cost and energy efficiency of inference at scale,” said Rogier Ketelaars, investment manager at Innovation Industries. “We believe the company is uniquely positioned to become a foundational player in the next generation of AI infrastructure.”

Find more reporting in our Technology section.