AI chip startup Axelera raises $68M to broaden offerings from edge to cloud

Edge artificial intelligence chip startup Axelera AI B.V., a Netherlands-based company that makes purpose-built silicon for generative AI and computer vision, today announced it has raised $68 million in recent funding.

The company’s Series B funding round is Europe’s largest oversubscribed Series B round within the fabless semiconductor industry and brings Axelera’s total funding thus far to $120 million.

Major institutional investors joining in the oversubscribed round included Invest-NL Deep Tech Fund, the European Innovation Council Fund, the Innovation Industries Strategic Partners Fund and the Samsung Catalyst Fund. Existing investors also participating included Verve Ventures, Innovation Industries, Fractionelera and the Italian sovereign fund CDP Enterprise Capital SGR.

Founded in 2021, Axelera AI is the developer of specialised computer chip hardware optimized to run AI software on the network edge, outside data centers. Its compute platform hardware accelerates AI inference, the processing that chatbots and computer vision do, more efficiently and closer to the purpose where the information is produced.

The startup believes that by putting its technology closer to the purpose where the AI processing must occur, it can speed up AI interactions. This will allow AI to have greater impact and control when run across connected devices similar to web of things systems and industrial robots.

“There’s no denying that the AI industry has the potential to rework a mess of sectors,” said Fabrizio Del Maffeo, co-founder and chief executive at Axelera AI. “Nonetheless, to actually harness the worth of AI, organizations need an answer that delivers high performance and efficiency while balancing cost.”

The company’s flagship product is the Metis AI Platform, a system based on the company’s chip, which is alleged to be able to 214 trillion computations per second, or 214 TOPS. In response to Axelera, the platform shouldn’t be just highly performant but is also very energy-efficient compared with traditional chips, making it highly suitable for edge devices. Many connected devices must run on battery for protracted periods.

Axelera said that it will use the funding to expand its sights towards broadening product offerings from the sting to cloud data centers to handle the needs of generative AI, large language models and multi-modal models. Not all machine learning and AI can run on the sting, since some processing requires heavy lifting in data centers.

“By expanding our product lines beyond the sting computing market, we are in a position to address industry challenges in AI inference and support current and future AI processing needs with our scalable, proven technology,” said Del Maffeo.

With this market expansion, Axelera will include high-performance computing by designing high-efficiency, high-performance but price-completive AI accelerators that might help when AI inference must scale out of the sting and must be migrated into the cloud. The corporate said that the recent chips will power future high-performance exascale and petascale data centers.

Image: Axelera