ºìÌÒÊÓÆµ

Innovating technologies for neuromorphic computing

23 - 25 November 2026 09:00 - 16:00 ºìÌÒÊÓÆµ Free Watch online

Discussion meeting and Transforming our Future event organised by Professor Stephen Furber CBE FREng FRS, Professor Judith Driscoll FREng FRS, Professor Stuart Parkin FREng FRS, Dr Sebastian C Dixon and Professor Tony Kenyon

The world is on the brink of an AI revolution. However, AI’s demand for computational power, doubling every 2–3 months, is unsustainable. Neuromorphic (brain-inspired) hardware offers transformative gains in energy efficiency without compromising performance. Achieving this vision requires close collaboration across materials, devices, systems and applications. To explore these themes, we are holding: 

  • A two-day Discussion Meeting (23–24 November), bringing together leading academic researchers to explore fundamental advances and challenges in neuromorphic computing, fostering new interactions and collaborations across disciplines
  • A one-day Transforming our Future event (25 November), focused on broader impacts, translation, and future directions for neuromorphic technologies

Programme

The programme, including speaker biographies and abstracts, is available below. Please note that the programme may be subject to change.

Poster session

There will be a poster session on Monday 23 November 2026. If you would like to present a poster, your proposed title, abstract (up to 200 words), author list, and the name of the proposed presenter and institution no later than Friday 23 October 2026. Acceptances may be made on a rolling basis so we recommend submitting as soon as possible in case the session becomes full. Submissions made within one month of the meeting may not be included in the programme booklet.

Attending the event

This event is intended for researchers in relevant fields.

  • Free to attend
  • Both virtual and in-person attendance is available. Advance registration is essential
  • Lunch is available on Day 1 and Day 2 of the meeting for an optional £25 per day. Lunch is available free of charge on Day 3. There are plenty of places to eat nearby if you would prefer to purchase food offsite. Participants are welcome to bring their own lunch to the meeting
  • Attendees are welcome to join for any of the meeting days; attendance across all three days is not required.

Please note that scientific meetings hosted by ºìÌÒÊÓÆµ do not necessarily represent a Royal Society position or signify an endorsement of the speakers or content presented.

Enquiries: contact the Scientific Programmes team

Image credit: © iStock.com / vchal

Organisers

  • Professor Judith Driscoll FREng FRS

    Professor Judith Driscoll FREng FRS

    Judith Driscoll is Professor in the Materials Science at the University of Cambridge and Fellow of Trinity College. She is Royal Academy of Engineering Chair in Emerging Technologies. She has been visiting faculty/staff member at Los Alamos National Lab. for 23 years. She is fellow of ºìÌÒÊÓÆµ of London, the Royal Academy of Engineering, APS, AAAS, MRS, RSC, IOP, IOM3, WES. Her research covers wide ranging oxide thin film engineering for electronic applications. She has many licensed patents, consults widely with industry worldwide, and has two start-ups. She is a strong supporter of women in science/engineering. She has numerous prizes for her work from Materials, Physics, Chemistry and EE societies.

  • Professor Stuart Parkin FREng FRS

    Professor Stuart Parkin FREng FRS

    Stuart Parkin’s research interests include atomically engineered thin-film heterostructures, high-temperature superconductors, and spintronic materials and devices for advanced sensor, memory and logic applications. Stuart’s discoveries in magnetoresistive thin film structures enabled a thousandfold increase in the storage capacity of magnetic disk drives.

    Most recently, Stuart’s research has focused on a novel storage class memory device, ‘racetrack memory’, and cognitive materials that could enable very low power computing technologies. He is a Member of the US National Academy of Sciences, the US National Academy of Engineering, the German National Academy of Sciences, Leopoldina, a Fellow of the American Academy of Arts and Sciences, and TWAS, (The World Academy of Sciences), and an Honorary Fellow of the Indian Academy of Sciences.

    Stuart has received numerous awards and honours, including, most recently, the 2012 Von Hippel Award from the Materials Research Society, the 2013 Swan Medal of the Institute of Physics, and the €1 million 2014 Millennium Technology Award from the Technology Academy Finland. He is also an Honorary Fellow of Trinity College, Cambridge.

  • Dr Sebastian Dixon

    Dr Sebastian C Dixon

    Seb Dixon is Network Manager for NeuMat (), a UK-wide initiative advancing neuromorphic materials and devices, and a postdoctoral researcher in the Department of Materials Science & Metallurgy at the University of Cambridge. His research focuses on pioneering non‑volatile memory materials to enable next-generation compute architectures, with particular emphasis on ferroelectric oxides, epitaxial growth, and the scalable integration of 2D materials for future compute. He leads a collaborative £1 million SMART grant between Cambridge and Paragraf Ltd, reflecting his strong track record in bridging academic research with industrial application and supported by his RAEng Industrial Fellowship. Seb’s international academic training spans London, Western Australia and Tokyo, complemented by an industrial EngD and seven years’ experience as a Materials Scientist in a UK semiconductor manufacturing start-up.

  • Professor Steve Furber CBE FREng FRS

    Professor Steve Furber CBE FREng FRS

    Steve Furber is a mathematician and engineer, revered for his fundamental contributions to electronic systems. He is known for his design of microprocessors, which control all electronic devices. His work has infinite applications, including controlling household appliances, medical devices such as pacemakers and military equipment.

    Steve was the original designer of the ARM processor, the world’s leading embedded processor core - bringing success to the United Kingdom in terms of engineering prestige and commercial value. His work assisted the recent explosion in wireless devices that use low-power microprocessors.

    His latest project is SpiNNaker - a computer that mimics the structure of the human brain. He is a Fellow of the Royal Academy of Engineering, the IEEE, the BCS and the IET, and has won numerous awards, including the IET Faraday Medal in 2007. In 2014, he became a Distinguished Fellow of the British Computer Society, joining Bill Gates, Tim Berners Lee and other preeminent pioneers of technology.

  • Image of Tony Kenyon

    Professor Tony Kenyon

Schedule

Chair

Dr Adnan Mehonic

Dr Adnan Mehonic

University College London, UK

09:00-09:05 Welcome by ºìÌÒÊÓÆµ and lead organiser
09:05-09:30 Neuromorphic computing in the Human Brain Project

An important aspect of the EU Flagship Human Brain Project (HBP: 2013-2023) was a focus on brain-inspired computing. In this context the HBP supported both the establishment of open neuromorphic computing services on the BrainScaleS wafer-scale analogue platform at Heidelberg and the SpiNNaker many-core platform at Manchester, and the development of second generation versions of both systems. Both platforms used the PyNN (Python Neural Network) language as their primary configuration tool and ran millions of jobs submitted remotely by hundreds of users around the world. The services continue today under the auspices of the EU EBRAINS 2.0 project.

Alongside these developments in neuromorphic computing the world has seen an explosion in the use and capabilities of AI. These AI systems are based principally on dense matrix operations that are supported on GPU hardware, However, this “compute everything all the time” approach has led to unsustainable energy requirements for AI. Biological brains point to a more sustainable approach where energy-consuming computation is sparse in both space and time. Could such an approach break the “deathly embrace” between current AI hardware and algorithms, and might this make neuromorphic computing platforms that support brain-like computation the platforms of choice for future energy-efficient AI systems?

Professor Steve Furber CBE FREng FRS

Professor Steve Furber CBE FREng FRS

University of Manchester

09:30-09:45 Discussion
09:45-10:15 Talk title TBC
10:15-10:30 Discussion
10:30-11:00 Break
11:00-11:30 Talk title TBC
11:30-11:45 Discussion
11:45-12:15 Talk title TBC
Professor Stuart Parkin FREng FRS

Professor Stuart Parkin FREng FRS

Max Planck Institute for Microstructure Physics, Germany

12:15-12:30 Discussion

Chair

Professor Judith Driscoll FREng FRS

Professor Judith Driscoll FREng FRS

University of Cambridge, UK

13:30-14:00 Neuromorphic electronics enabled by 2D materials

As neuromorphic computing moves from laboratory demonstrations toward real-world deployment, reliability and environmental robustness have become increasingly important challenges. In this talk, I will discuss how atomically thin two-dimensional (2D) materials and van der Waals (vdW) interfaces can provide new opportunities for building reliable and energy-efficient neuromorphic hardware.

Unlike conventional bulk materials, vdW materials offer precisely engineered interfaces and atomically controlled transport pathways that enable improved device uniformity, stability, and functionality. I will present our recent work on memristive devices in which a 2D interfacial layer suppresses unwanted ionic diffusion and enables robust analog switching at elevated temperatures. I will also introduce a universal tunnel selector based on multiple vdW layers that exhibits exceptionally low cycle-to-cycle variation, minimal temperature dependence, and record-high endurance and nonlinearity. Together, these advances establish key building blocks for large-scale neuromorphic systems with improved reliability and manufacturability.

Beyond memory and selector devices, vdW materials provide a versatile platform for engineering ionic and electronic dynamics that emulate learning, adaptation, and temporal information processing in biological neural systems. By combining these neuromorphic functionalities with the intrinsic resilience of engineered 2D interfaces, we establish pathways toward intelligent hardware capable of operating in harsh conditions, including high-temperature industrial settings, energy infrastructure, and future space systems.

These results highlight the emerging role of 2D materials in neuromorphic computing, where atomic-scale control of interfaces can simultaneously enable efficient intelligence, device reliability, and operation in extreme environments.

Dr Joshua Yang

Dr Joshua Yang

University of Southern California, US

14:00-14:15 Discussion
14:15-14:45 Memristors and future energy-efficient computing systems

Emerging AI workloads are placing increasing pressure on the energy efficiency of today’s computing systems. This talk will explore how memristive devices can enable new memory and computing architectures that significantly reduce data movement and power consumption. I will highlight recent progress in device technology and discuss the path towards scalable, industry-relevant implementations.

Dr Adnan Mehonic

Dr Adnan Mehonic

University College London, UK

14:45-15:00 Discussion
15:00-15:30 Break
15:30-16:00 Electrochemical ionic synapses for energy-efficient brain-inspired computing

In this talk, I will share our work on the ionic electrochemical synapses, whose electronic conductivity we control deterministically by electrochemical insertion/extraction of dopant ions into/out of the channel layer. This work is motivated by the need to enable significant reductions in the energy consumption of computing, and is inspired by the ionic processes in the brain. Proton as the working ion in our research presents with very low energy consumption, on par with biological synapses in the brain. Our modelling results indicate the desirable material properties, such as ion conductivity and interface charge transfer kinetics, that we must achieve for fast (ns), low energy (< fJ) and low voltage (<1V) performance of these devices. Importantly, the target material is a mixed proton-electron conductor, whose electronic conductivity depends on the proton concentration through doping and phase change effects. The candidate materials are a spectrum of intercalation oxides as well as 2D van der Waals materials. We have assessed the electron polaron and proton mobilities in these systems, to understand the conductivity modulation mechanisms, and down-select most promising materials. In addition, the conductance change in these electrochemical devices depends non-linearly on the gate voltage, due to field-enhanced ion migration in the electrolyte, and charge transfer kinetics at the electrolyte-channel interface. We are leveraging these intrinsic nonlinearities to emulate bio-realistic learning rules deduced from neuroscience studies, such as spike timing dependence of plasticity and Hebbian learning rules. Our findings provide pathways towards brain-inspired hardware that has high yield and consistency and uses significantly lesser energy as compared to current computing architectures.

Professor Bilge Yildiz, Massachusetts Institute of Technology, USA

Professor Bilge Yildiz, Massachusetts Institute of Technology, USA

16:00-16:15 Discussion
16:15-16:45 Talk title TBC
Professor Harish Bhaskaran FREng

Professor Harish Bhaskaran FREng

University of Oxford, UK

16:45-17:00 Discussion

Chair

Dr Sebastian Dixon

Dr Sebastian C Dixon

Neumat

09:00-09:30 Radio-frequency spintronic neural networks

Radio-frequency spintronic neural networks offer several distinctive advantages over conventional in-memory computing approaches.

First, they naturally operate in the frequency domain, with frequency-multiplexed inputs and propagating signals throughout the network. This enables fully parallel inference while greatly simplifying the wiring architecture.

Second, they make it possible to reconfigure all synaptic weights remotely, without selector devices or individual access lines. Programming signals are frequency-multiplexed in the same way as inference signals and are delivered through the same access paths, which simplifies programming, opens the path to shared circuitry for programming and inference, and promises reduced footprint together with three-dimensional integration.

Third, these networks can reconfigure not only synaptic weights but also network connectivity during learning. In such systems, synapses and neurons interact when their frequencies match: small frequency shifts primarily modify the synaptic weight, whereas larger shifts can reroute a synapse toward a different neuron. I will show that magneto-ionic effects can tune synaptic frequencies over a sufficiently wide range to modify the network topology itself, enabling on-chip learning with a drastically reduced number of devices compared with fixed-topology networks.

Dr Julie Grollier

Dr Julie Grollier

Université Paris-Saclay, France

09:30-09:45 Discussion
09:45-10:15 Talk title TBC
10:15-10:30 Discussion
10:30-11:00 Break
11:00-11:30 Precision engineered WO3 hybrid switching memristors providing low voltage, uniform analogue switching for inference accelerators

Opportunities for oxide-based resistive-switching devices in edge artificial intelligence (AI) inference accelerators are significant, as they promise energy-efficient and low-latency hardware. However, practical deployment is constrained by materials challenges. We  address these challenges through precision engineered, CMOS compatible WO3-x thin film memristor devices based on WO3-x/TiN made at low temperature. We achieve hybrid (filament+interface) switching at low current.  Interface switching is enabled at low voltage and stable binary resistive switching with endurance exceeding 10³ direct-current (DC) cycles and 10âµ pulsed cycles are achieved with non-volatile retention beyond 10â´ s, confirming a robust and reliable switching landscape which combines advantages of both the filament and interface approaches. Building on this stability, identical low-voltage programming pulses (±0.5 V) enable linear, well-separated multi-level conductance states with high cycle-to-cycle (C2C) and device-to-device (D2D) uniformity, are achieved indicating controlled analogue modulation. Overall, this fully CMOS-compatible and manufacturable device architecture provides a scalable pathway toward reliable analogue AI inference accelerators.

Professor Judith Driscoll FREng FRS

Professor Judith Driscoll FREng FRS

University of Cambridge, UK

11:30-11:45 Discussion
11:45-12:15 Talk title TBC
Professor Daniele Ielmini

Professor Daniele Ielmini

Politecnico di Milano, Italy

12:15-12:30 Discussion

Chair

Professor Stuart Parkin FREng FRS

Professor Stuart Parkin FREng FRS

Max Planck Institute for Microstructure Physics, Germany

13:30-14:00 Talk title TBC
14:00-14:15 Discussion
14:15-14:45 Ferroelectric field-effects with hafnium oxide for neuromorphic hardware

In ferroelectric resistive weights, the strength of the synaptic connection between two neurons is stored in the device conductance. During learning, programming pulses are applied to the synaptic weight, which reconfigures the ferroelectric domains and adjusts the conductance. One strategy to lower the energy cost during the training phase is to lower the duration of the programming pulses. However, the latter cannot be shorter than the self-loading time of the resistive weights, limited by parasitic delays in the circuits. We fabricate ferroelectric resistive weights using bilayers based on hafnia/zirconia superlattices and tungsten oxide. Using this process, CMOS Back-End-Of-Line integration was demonstrated. We determine the maximal device area for which the self-loading time becomes sufficiently short to enable 20 ns programming, which corresponds to a maximum of 3 pJ per pulse. We show that spiking neural network can be deployed on these devices for adaptive electroencephalography decoding. Finally, ferroelectric capacitors based on the same material also exhibit low-power programming and fast switching speed: full ferroelectric domain reversal is obtained for 5V pulses of only 1 ns.

Dr Laura Bégon-Lours

Dr Laura Bégon-Lours

ETH Zürich, Switzerland

14:45-15:00 Discussion
15:00-15:30 Break
15:30-16:00 Molecular neuromorphic building blocks for artificial intelligence

Artificial Intelligence has long oscillated between grand promises and inevitable disillusionment. Remarkable milestones, such as, AI outperforming human champions in complex games, generating fluent language, diagnosing disease suggest we are entering a new era, yet a deeper look reveals that these breakthroughs come at a steep cost: vast energy consumption and intensive, expensive training. In cognition, decision-making, and adaptive intelligence, even the most advanced computing machines fall far short of the brain’s efficiency and compact design. The core challenge lies in the limitations of conventional circuit elements and architectures, which struggle to replicate the brain’s nonlinear dynamics. In this talk I will introduce a new class of molecular circuit elements designed to capture reconfigurable, brain-like logic at the nanoscale. These devices operate as analog or digital elements, or can be poised at the edge of instability, offering a unique capacity to emulate neural functions in ways that traditional hardware cannot. I will trace the journey from the foundational physics and chemistry of these molecular systems through to integrated circuit design and on-chip applications, with the aim of laying the groundwork for AI platforms that transcend the limitations of Moore’s Law and open a new era of energy-efficient computing.

Professor Sreetosh Goswami

Professor Sreetosh Goswami

Indian Institute of Science, India

16:00-16:15 Discussion
16:15-17:00 Panel discussion - future directions

Chair

Professor Martin Dawson FRS

Professor Martin Dawson FRS

University of Strathclyde

09:30-09:40 Welcome and opening remarks
Professor Martin Dawson FRS

Professor Martin Dawson FRS

University of Strathclyde

09:40-10:20 Keynote
Mike Davies

Mike Davies

Intel

Chair

Professor Judith Driscoll FREng FRS

Professor Judith Driscoll FREng FRS

University of Cambridge, UK

10:20-10:25 Introduction from the Chair
Professor Judith Driscoll FREng FRS

Professor Judith Driscoll FREng FRS

University of Cambridge, UK

10:25-10:50
Dr Markus Hellenbrand

Dr Markus Hellenbrand

University of Cambridge, Intrinsic Semiconductor Technologies

10:50-11:15
Professor Bipin Rajendran

Professor Bipin Rajendran

King's College London, UK

Chair

Image of Tony Kenyon

Professor Tony Kenyon

University College London

11:45-11:50 Introduction from the Chair
Professor Tony Kenyon

Professor Tony Kenyon

University College London

11:50-12:15
Dr Heba Bevan

Dr Heba Bevan

UtterBerry Limited

12:15-12:40
Dr Bradley Theilman

Dr Bradley Theilman

Sandia National Laboratories

Chair

Professor Steve Furber CBE FREng FRS

Professor Steve Furber CBE FREng FRS

University of Manchester

13:40-13:45 Introduction from the Chair
Professor Steve Furber CBE FREng FRS

Professor Steve Furber CBE FREng FRS

University of Manchester

13:45-14:10
Professor André van Schaik

Professor André van Schaik

University of Manchester

14:10-14:35
Dr Hector Gonzalez

Dr Hector Gonzalez

SpiNNcloud

14:35-15:05 Panel: The future of neuromorphic computing
Professor Judith Driscoll FREng FRS

Professor Judith Driscoll FREng FRS

University of Cambridge, UK

Professor Steve Furber CBE FREng FRS

Professor Steve Furber CBE FREng FRS

University of Manchester

Professor Tony Kenyon

Professor Tony Kenyon

University College London

Professor Sergei Turitsyn

Professor Sergei Turitsyn

Aston University

15:05-15:15 Closing remarks
Professor Martin Dawson FRS

Professor Martin Dawson FRS

University of Strathclyde