1st Minor Meeting: NeuMat Kick-Off Meeting

Shamit Shrivastava, Apoha.aiI

Liquid Brain: From Excitable Substrates to Sensory Processors

We present Liquid Brain®, a neuromorphic platform built on the invention of synthetic neurons — engineered excitable substrates that reproduce the physics of sensation. By transforming contact into waves, waves into signals, and signals into high-dimensional state diagrams, Liquid Brain creates a new dataclass for computation at the material level. This represents a shift from simulating neural architectures to constructing neuromorphic materials, where sensing and computation are inseparably coupled.
Our results show that synthetic neurons enable rich, nonlinear, and multiparametric encoding of molecular interactions from microgram-scale samples, demonstrating neuromorphic substrates as practical sensory processors. This establishes a pathway for neuromorphic systems to capture behavior that conventional instruments cannot. Alongside this invention, we discuss a benchmarking perspective: aligning evaluation with market-driven, multiparametric tasks rather than synthetic abstractions. We argue that such benchmarks can accelerate adoption and scaling, bridging the gap from foundational physics to deployable neuromorphic technologies.

J. Joshua Yang University of Southern California, Los Angeles, CA 90089, USA – Email: jjoshuay@usc.edu

Neuromorphic Computing via Integration of Diffusive and Drift Memristors

Memristors can be broadly categorized into diffusive and drift types, distinguished by their reset switching mechanisms. Diffusive memristors reset through spontaneous ion diffusion under zero bias, exhibiting dynamics that closely resemble biological ionic processes, making them particularly attractive for processing temporal information. Drift memristors, in contrast, rely on electric-field-driven ion migration, offering highly stable and tunable analog resistance states that are well suited for storing long-term information. The synergistic integration of diffusive and drift memristors enables the design of compact, energy-efficient, and biorealistic neuromorphic hardware. This talk will present recent progress in device engineering, array-level integration, and application demonstrations, highlighting their potential to advance next-generation computing architectures.
J. Joshua Yang is a professor of the Department of Electrical and Computer Engineering at the University of Southern California. His current research interest is Post-CMOS hardware for neuromorphic computing, machine learning and artificial intelligence, where he published several pioneering papers and holds 120 granted US Patents. He is the Founding Chair of the IEEE Neuromorphic Computing Technical Committee and the director of USC-Airforce Center of Excellent on Neuromorphic Computing. He was elected to the IEEE Fellow and the National Academy of Inventors (NAI) Fellow for his contributions to resistive switching materials and devices for nonvolatile memory and neuromorphic computing.

Sergei K. Turitsyn, Aston institute of Photonic Technologies, Aston University

UK Multidisciplinary Centre for Neuromorphic Computing

This presentation will introduce the recently established UK Multidisciplinary Centre for Neuromorphic Computing. Funded by UKRI–EPSRC, the Centre aims to identify and address fundamental research challenges in neuromorphic computing. It will serve as a focal point for the UK research community, fostering collaborations across disciplines and internationally. We will outline the Centre’s research agenda and planned activities to engage with researchers across relevant disciplines, industry, and policymakers.

Giuliana Di Martino, Cambridge University

Bridging Disciplines to Advance Neuromorphic Technologies

The development of next-generation neuromorphic systems demands tight integration across traditionally separate domains—materials science, device engineering, and algorithm design. Yet, meaningful progress requires more than just collaboration; it calls for a shared understanding of the challenges and priorities at every level of the stack.
In this talk, I will reflect on my own research journey as an example of the multidisciplinary approach needed to make impactful advances in neuromorphic technologies. My work spans from characterising advanced functional materials to building and evaluating devices under realistic operational conditions. This requires continuous dialogue and co-design across disciplines: materials must be engineered with specific device architectures in mind, devices must be interrogated both structurally and functionally, and all of this must be informed by the performance needs of higher-level computing paradigms. Through examples ranging from memristors and 2D materials to ferroelectrics and magnetic systems [1-5], I will illustrate how nano-optical characterisation techniques such as ultrafast spectroscopy, near- field microscopy, and single-particle imaging can provide crucial insights into material dynamics—such as defect migration and phase transitions—that directly impact device performance. I will also share how integrating this knowledge into the design of neuromorphic components has raised important research questions that could not be formulated without a clear view across the stack. This is precisely the type of challenge the NeuMat network seeks to address: creating a connected, collaborative research community able to identify the most meaningful problems and work collectively towards energy-efficient, high-performance neuromorphic systems.

1. Nature Electronics (2020), 3, 687 – G. Di Martino, A. Demetriadou, W. Li, D. Kos, B. Zhu, X. Wang, B. de Nijs, H. Wang, J. MacManus-Driscoll and J. J. Baumberg
2. Advanced Materials (2022), 2209968 – J. Symonowicz, D. Polyushkin, T. Mueller, G.
3. Small (2025), 2410569 - D. Kelly, J. Symonowicz, C. Stewart, S. Hoffman, G. Di Martino
4. Advanced Functional Materials (2023), 2214970 – A. Jan, T. Rembert, S. Taper, J. Symonowicz, Nives Strkalj, T. Moon, Y. Seong Lee, H. Bae, H. Jae Lee, D. Choe, J. Heo, J. MacManus-Driscoll, B. Monserrat, G. Di Martino
5. Small Science (2024), 2400223 - A. Jan, S. Fraser, T. Moon, Y. Seong Lee, H. Bae, H. Jae Lee, D. Choe, M. T. Becker, J.L. MacManus-Driscoll, J. Heo, G. Di Martino

Abu Sebastian, IBM

Bridging the hardware divide between biological and artificial neural networks?

Deep artificial neural networks or deep learning, which has revolutionized AI in recent years, is algorithmically rooted in computational neuroscience. However, its hardware implementation was never neuro-inspired. This mismatch created a “hardware divide,” resulting in significant energy inefficiency. Much of today’s research in classical computing is focused on closing this divide. Over the past decade, AI computing systems have gradually adopted brain-inspired features. Modern custom digital deep learning accelerators leverage characteristics such as massive parallelism and reduced-precision arithmetic. In-memory computing accelerators, which are rapidly maturing, incorporate the attribute of in-place processing. In this talk, I would like to share my perspective on what further innovations and breakthroughs are needed to narrow the “hardware divide.”

Paul Larcey, Innovate

From the lab to the market: How Innovate UK can help the journey:

University spin-offs that emerge from academic research often encounter significant challenges when transitioning from low Technology Readiness Levels (TRLs 1-4) to propositions ready for investors. This process requires systematic development in technical validation, intellectual property strategy, market analysis, business model creation, and team formation to bridge the vital "valley of death" between research funding and commercial investment. Neuromorphic computing will provide crucial technologies for the full optimisation of AI, robotics, and other emerging sectors that demand energy-efficient computing over the coming years. Innovate UK Business Connect plays an essential role in supporting the path from laboratory to market through its comprehensive network of services. These include business development assistance such as access to experienced mentors and advisors, extensive networks connecting to potential customers, partners, and investors, structured programmes for creating investor-ready business plans and financial forecasts, workshops on pitching and intellectual property commercialisation, and tailored one-to-one support. Innovate UK's broader funding landscape offers vital bridging finance through grants and loans, enabling spin-offs to reach higher TRL levels and reduce risks before engaging with private investors.

Tony Kenyon, UCL

Neuroware – the UK Neuromorphic Computing Hardware Semiconductor Innovation and Knowledge Centre (IKC)

1^st October 2025 sees the start of the UK Neuromorphic Computing Hardware Semiconductor Innovation and Knowledge Centre (IKC), which we have named Neuroware. This is a multi-institution and multi-partner centre that aims to create real world impact from developing brain-inspired, neuromorphic, technologies. Drawing on fundamental research from our academic core partners – UCL, Cambridge, Oxford, Imperial, King’s College London, Manchester, Sheffield, Strathclyde and NPL, along with 30+ industry partners, the centre will accelerate technologies to higher technology readiness levels (TRLs) to maximise their potential for commercialisation.
Neuroware is deliberately broad in its remit, with three main technology strands – CMOS-based systems, new and emerging technologies, and neuromorphic photonics – underpinned by a solid base of computational neuroscience. The centre will also be highly inclusive and open to further collaboration through a series of workshops, sandpits and calls for funding of pump-priming activities.
I will describe Neuroware’s mission and structure, highlighting how it fits within the emerging UK
neuromorphic ecosystem and plugs in to global initiatives.

Professor Themis Prodromakis, Regius Chair of Engineering, Centre for Electronics Frontiers, Institute for Micro Nano Systems, University of Edinburgh, Edinburgh, EH9 3FB, UK - Email: t.prodromakis@ed.ac.uk

Innovations across AI and Semiconductors

The 21^st century is defined by increasingly intelligent machines and a drive to use them for augmenting human capability. On one side, developments in Artificial Intelligence (AI) are more and more inspired by nature’s efficiency and capitalize on innovations in semiconductor technologies. On the other hand, advances in electronics are increasingly assisted through the use of AI tools and capabilities that accelerate the design, manufacturing, modelling and testing cycles of next generation electronics and semiconductor technologies. In this talk Professor Prodromakis will present a few examples across the AI and Semiconductor remits, including memristor-based AI hardware and the latest work of the APRIL AI Hub.
Biography: Themis holds the Regius Chair of Engineering at the University of Edinburgh and is Director of the Centre for Electronics Frontiers. His work focuses on developing energy-efficient AI hardware solutions through innovating novel semiconductor technologies and neuromorphic computing architectures. He leads an interdisciplinary team comprising researchers with expertise across materials process development to electron devices and circuits and systems for applications in embedded systems and AI. He holds an RAEng Chair in Emerging Technologies and is Adjunct Professor at UTS Australia and Honorary Fellow at Imperial College London. He is Fellow of the British Computer Society, the Royal Society of Chemistry, the IET and the Institute of Physics. He is also the Director of the UKRI APRIL AI Hub that is developing AI tools and capabilities for the electronics sector. In 2015, he established ArC Instruments Ltd that delivers high-performance testing infrastructure for automating characterisation of novel nanodevices in over 26 countries. His contributions in memristive technologies and applications have brought this emerging technology one step closer to the electronics industry for which he was recognised as a 2021 Blavatnik Award UK Honoree in Physical Sciences and Engineering and with the 2025 Princess Royal Silver Medal from the Royal Academy of Engineering.