Axelera AI Secures $250 Million to Scale Energy-Efficient Chips for Edge Computing

Axelera AI Closes $250 Million Funding Round Led by BlackRock and Innovation Industries

Axelera AI, a Dutch European AI chip startup based in Eindhoven, has secured more than $250 million in one of the largest investments to date in a European AI chipmaking firm 1

1

2

2

. The funding round was led by European venture capital fund Innovation Industries, with participation from new investors including BlackRock and SiteGround Capital, alongside existing backers such as Samsung Catalyst Fund, Bitfury, Verve Investments, and funds backed by Belgian and Dutch governments 2

2

.

Since incorporating in 2021, Axelera AI has attracted over $450 million in equity, grants and venture debt, positioning itself as a significant player in the competitive AI chip landscape 1

1

. The company's customer base has more than tripled in the past year, reflecting growing demand for alternatives to traditional GPU-based infrastructure 1

1

.

Targeting Edge AI with Power-Efficient Semiconductors

Axelera AI specializes in manufacturing AI chips for edge devices that run inference workloads—the process of deploying AI models in real-world applications after training 1

1

. Unlike training, which typically occurs in large data centers using powerful GPUs like those from Nvidia, inference can be deployed locally on devices such as smartphones, security cameras, or warehouse robots 3

3

. The company's energy-efficient inference chips are designed specifically for these edge applications, aiming to reduce reliance on costly, power-hungry data center infrastructure 2

2

.

Chief Executive Officer Fabrizio Del Maffeo explained that customers are actively seeking more accessible compute solutions, improved efficiency, and ways to reduce spending on expensive AI infrastructure 1

1

. He emphasized that inference can run on a decentralized chip architecture that consumes significantly less power, addressing concerns about straining power grids and rising energy costs associated with massive data centers 1

1

.

Digital In-Memory Computing Architecture Powers Metis AI Chip

Axelera AI's flagship product, the Metis AI chip, can perform 214 trillion computations per second while consuming approximately 10 watts of power 3

3

. This low power consumption makes it particularly suitable for battery-powered connected devices operating in industrial settings. The chip is powered by a proprietary architecture called digital in-memory computing, or D-IMC, which dramatically reduces power usage by limiting data movement within the processor 3

3

.

Traditional AI processors store and process data using separate circuit sets, requiring constant data shuffling that consumes substantial power. Axelera AI's D-IMC architecture uses SRAM, a high-speed memory variety, to store and process information in the same location, organized in a crossbar array configuration that can carry out calculations on vectors and matrices—the fundamental data units used by AI models 3

3

. The company ships Metis as part of two accelerator cards that attach to connected devices via PCIe or M.2 interfaces, and provides customers with the Voyager SDK software toolkit and a Model Zoo of pre-packaged AI models .

Europa Chip Launch and Expanding Market Opportunities

The $250 million funding round will be used to expand manufacturing of Axelera AI's next-generation Europa chip, which the company plans to launch before June 2026, and to develop software that simplifies chip integration for customers 2

2

. The Europa chip represents a significant performance leap, capable of performing 629 trillion computations per second—more than twice the capacity of Metis 3

3

.

Europa includes 8 AI-optimized cores, 16 central processing unit cores, and 128 megabytes of memory, delivering up to three times the performance per watt compared to competing products 3

3

. The chip is particularly suited for running computer vision models, with internal testing showing it can process more than 13,168 frames per second 3

3

. Del Maffeo noted that Axelera has "a large pipeline of opportunities that require investments and this pipeline is increasing" 1

1

.

Additionally, Axelera AI received a $66 million grant in March 2025 as part of a European Union project to develop an advanced chip called Titania for use in supercomputing centers, also referred to as AI factories, expected to launch in 2027 2

2

. The inference market is expanding rapidly as companies seek to monetize their AI investments, with competitors like Nevada-based Positron raising $230 million this month for similar AI chips for inference applications 1

1

.