Dresden-based company Semron has successfully raised $7.9 million in seed funding for its AI chips featuring 3D packaging designed for mobile devices. They aim to set a new standard in AI chip technology within the mobile industry.
Semron’s funding was led by Join Capital and backed by SquareOne, OTB Ventures and Onsight Ventures.
Jan Borgstädt, the managing partner at Join Capital, says: “We recognize the immense potential of Semron’s innovation to redefine the landscape of AI capabilities in smart devices globally and to start a new dynamic in the semiconductor industry that is struggling with the era of GenAI.”
Semron aims to develop advanced computer chips to fuel the growing demand for sophisticated AI capabilities. These chips are specifically designed to power small consumer devices such as smartphones, earbuds and VR headsets.
The company focuses on markets requiring strong AI, specifically edge computing devices. These include smartphones with large language models (LLMs)/GenAI for digital assistants and video processing, earbuds for voice control and live translation and VR headsets with LLMs/GenAI.
Semron plans to use the capital infusion to enhance hardware and compiler development, expand its team and prioritize internationalization. The company currently has eleven employees and plans to quadruple the team size by year-end.
Semron notes that its 3D semiconductor technology achieves up to 20x chip efficiency, enabling the execution of up to 1000x larger AI models without expanding the chip size.
Meanwhile, Semron’s technology, CapRAM, uses a semiconductor device architecture based on a variable capacitor called a ‘memcapacitor.’
Aron Kirschen, CEO of Semron, says the company’s technology sets it apart from competitors like Nvidia. Their focus on performance, cost-efficiency and targeting the edge computing market for advanced AI features gives them an advantage, Kirshen asserts.
“Our advantage lies in our technology’s ability to run several hundred times larger AI models and to deliver up to 20 times the efficiency at a cost two orders of magnitude lower per unit of performance,” says Kirshen. “This is achieved through the paradigm shift to stack many compute layers on the same silicon substrate.”
To date, the company says it has raised $10.9 million.
Startup EnCharge AI receives $21.7M to develop advanced edge AI chip tech
Mythic secures $13M for edge AI chip; Lockheed Martin Ventures among investors