Spin textures and spin waves for signal processing and sensor technologies

In the post-ITRS Roadmap era, beyond Moore's Law, traditional complementary metal-oxide-semiconductor (CMOS) technology is approaching its physical limits. Transistors have nearly reached their ultimate scaling potential, as evident in challenges such as high power losses in 3D integration and the increasing complexity of wiring nanoscopic components. Digital algebra with ferromagnetic multilayers has emerged as a promising avenue in Beyond-CMOS device research. Specifically, its implementation in *Perpendicular Nanomagnetic Logic* (pNML) offers unique advantages, including inherently non-volatile computing states, ultra-low energy consumption (attojoule losses per bit operation), and CMOS-compatible data throughput.

This technology enables system-level benefits such as monolithic 3D integration, field-based coupling for computational operations instead of electrical connections, and extensive parallelism via a synchronous magnetic field clock. Nanomagnetic components capitalize on the intrinsic properties of magnetic materials, such as the non-volatility of digital states, low energy requirements, radiation resistance, and cost-effective production using back-end-of-line processes like PVD and sputtering. These features make nanomagnetic components particularly well-suited for demanding applications in aerospace, automotive, and robotics industries, where they excel under high temperatures, radiation exposure, and harsh environmental conditions.

Another approach to magnetic data and signal processing involves spin-wave devices, which can be fabricated through direct structuring with a focused ion beam. Inspired by optical components such as lenses, gratings, and Fourier primitives, these devices enable the development of complex structures for edge AI implementations and, more broadly, programmable matter for unconventional, hardware-based computing in the physical domain. With their inherent parallelism, scalability, and high temporal dynamics (reaching gigahertz frequencies), spin-wave devices are promising candidates for implementing neuromorphic systems. These advancements have the potential to significantly enhance machine learning, artificial intelligence, and pattern recognition, while also enabling signal processing modeled on the human brain.

Please, enjoy, browsing our current research topics.