“At Nellow, we believe the future of AI depends on fundamentally rethinking how information is stored and processed. By merging spintronics and ferroelectricity, we’re unlocking computing architectures that are both radically energy-efficient and inherently scalable. Partnering with the EDGE AI ecosystem allows us to bring this technology to real-world applications—from datacenters to the smallest edge devices—faster than ever.”
— Jean-Philippe, CEO of Nellow
NELLOW, a European deep-tech startup developing a new class of Edge AI chips, is pioneering a groundbreaking class of technology that merges spintronics and ferroelectricity, creating computing components capable of both storing and processing information without continuous power consumption through a specific HW/SW co-design architecture.
These components overcome a fundamental limitation of today’s CMOS architectures by enabling non-volatile memory and logic information processing, with no power required to retain data. Best of all, the technology is fully compatible with standard CMOS fabrication, ensuring cost-effective manufacturing while delivering a step-change in energy efficiency, density, and performance.
While this innovation has major implications for High-Performance Computing (HPC) and AI acceleration in data centers, its potential at the edge is equally transformative. NELLOW’s technology enables a new generation of energy-efficient Edge AI products, from smart sensors and autonomous drones to intelligent industrial systems capable of local learning and inference at ultra-low power.
This partnership reinforces NELLOW’s mission to deliver ultra-efficient, secure, and multimodal AI processing including LLMs directly on-device: a radically new generation of Edge AI processor, built for local intelligence overcoming the limits of conventional CMOS.
