Plasticity in NEUral Memristive Architectures

Lead Research Organisation: University of Southampton
Department Name: Electronics and Computer Science


During the past two decades, philosophers, psychologists, cognitive scientists, clinicians and neuroscientists strived to provide authoritative definitions of consciousness within a neurobiological framework. Engineers have more recently joined this quest by developing neuromorphic VLSI circuits for emulating biological functions. Yet, to date artificial systems have not been able to faithfully recreate natural attributes such as true processing locality (memory and computation) and complexity (10^10 synapses per cm2), preventing the achievement of a long-term goal: the creation of autonomous cognitive systems.

This project aspires to develop experimental platforms capable of perceiving, learning and adapting to stimuli by leveraging on the latest developments of five leading European institutions in neuroscience, nanotechnology, modeling and circuit design. The non-linear dynamics as well as the plasticity of the newly discovered memristor are shown to support Spike-based- and Spike-Timing-Dependent-Plasticity (STDP), making this extremely compact device an excellent candidate for realizing large-scale self-adaptive circuits; a step towards "autonomous cognitive systems". The intrinsic properties of real neurons and synapses as well as their organization in forming neural circuits will be exploited for optimising CMOS-based neurons, memristive grids and the integration of the two into realtime biophysically realistic neuromorphic systems. Finally, the platforms would be tested with conventional as well as abstract methods to evaluate the technology and its autonomous capacity.

Planned Impact

The work proposed here will provide answers to an extremely diverse audience including: computer science, electronics, physics, neuroscience, medicine, education and many other communities that are intrigued by the way the human brain functions. Thus the anticipated impact will expand across three main areas:
1. Economy
2. Host Institutions (CBiT, IMSE, INI, IGI and LAAS), EU science-base
3. Society
Economic Impact: The main beneficiaries stand to be the semiconductor and micro/nanoelectronics industries. The proposed research promotes the use of memristors in complex networks that could be independently trained to solve tasks in which conventional supercomputers prove to be inadequate. The large impact of this newly discovered device is expected to be reflected in the semiconductor industry by adjusting processing techniques and materials to accommodate the fabrication of memristors along with conventional devices in standard technologies.
Memristive networks could potentially be mingled with existing processing architectures, as extra processing cores, for carrying out assignments in which perception, reasoning and motivation are required. We therefore anticipate that this research will have a significant impact in the computer-science industry, since this technological advancement will reshape the current computation strategies and it will eventually act as the enabler to new applications. Particular benefits are thus foreseen in leading companies that impose the designs of novel processing unit architectures like IBM, Intel and AMD. HP has already invested a lot in the development of memristors and memristive networks, demonstrating existing potential for future emerging technologies.
Although, digital computers are good in number crunching, they struggle with tasks like face recognition, real-time navigation control, object segmentation and depth perception. Therefore, it is not unrealistic to say that this project is a step towards the development of autonomous systems, with all the apparent benefits for the robotics industry.
An added benefit of the proposed technology is its extreme versatility, since the processing of different impulses is contingent on the training of the system. Memristive networks could also be applied to solve mathematical problems that do not formulate a specific algorithmic solution, with particular applicability in modeling stock trading. Last but not least, the proposed technology could be utilized to demonstrate associative indexing of information, which is particularly useful in large database architectures and the Internet for increasing the speed of information retrieval.

Impact to the Host Organisations and to the European Union: See Academic Beneficiaries.

Impact to Society: Biology makes excellent use of resources to solve a given task, being efficient, robust, adaptable, real-time, effective, scalable and reliable. For the above reasons,
any engineer would do extremely well in learning from nature. The experimental platforms we propose to develop would shine more light in the way the human brain, and biology in
general, learns and adapts according to its environment. This question has often been addressed by a broad range of disciplines, from philosophy to medicine and more recently
engineering. Thereby, one can appreciate how important it is to have the physical means of studying and understanding how knowledge evolves.
Apart from providing knowledge to society, we believe that this project could be further exploited, leading into advances on international development, primarily in healthcare. Without any doubt, stimulating patterns for treating epilepsy and other brain-related diseases could be safely tested on benchmarking memristive platforms, before initiating clinical trials. Finally, the "quality of life" enhancement is highlighted by the employment of synaptic-like architectures both in healthcare applications as well as automation, for example cruise control


10 25 50

Related Projects

Project Reference Relationship Related To Start End Award Value
EP/J00801X/1 04/09/2011 13/04/2013 £482,747
EP/J00801X/2 Transfer EP/J00801X/1 14/04/2013 04/03/2015 £218,244
Description To date, we have succeeded to realistically emulate synaptic plasticity with memristive nanodevices. This is the prime objective of PNEUMA as it serves as proof that memristors can indeed be employed as fundamental blocks for establishing unconventional (neuromorphic) computation hardware. To achieve this, we first reviewed the neuro- and neuromorphic computing approaches that can best exploit the properties of memristor and nanoscale devices, and then devised a novel hybrid memristor-CMOS neuromorphic circuit that represents a radical departure from conventional neuro-computing approaches, as it uses memristors to directly emulate the biophysics and temporal dynamics of real synapses. We have pointed out the differences between the use of memristors in conventional neuro-computing architectures and the hybrid memristor-CMOS circuit proposed, and argued how this circuit represents an ideal building block for implementing brain-inspired probabilistic computing paradigms that are robust to variability and fault tolerant by design.
Exploitation Route The work funded by PNEUMA has attracted world-wide interest, particularly Prof Leon Chua, who is the original proposer of memristors. This is also ascertained by Prof Chua spending three extended visits within Dr Prodromakis's group, including a two-year Marie Currie Fellowship starting in 2013 and more recently a Diamond Jubilee Visiting Fellowship (3 years) with the University of Southampton.

Other significances include further funding from EU-FP7 (RAMP €2.1M, ICT-2013-10: 612058) for exploiting the biorealistic attributes of memristors in establishing biomimetic neuronal bridges and advanced sensing topologies. This technology is anticipated be impactful in emerging neural interfaces. Additional funding was also secured by the EPSRC (Reliably Unreliable Nanotechnologies, £1.4M, EP/K017829/1) towards substantiating new design paradigms that can facilitate reliable computation and processing with inherently unreliable nanoscale elements. This research is particularly timely as we are currently reaching the scaling limits of conventional semiconductors.

Moreover, our findings have been widely disseminated via: > 15 invited talks (2012-2014), 1 Nature Materials Commentary and Dr Prodromakis's selection to chair CAS-FEST 2014 on the field of Memristors.
These activities have enabled: 1) the wide use of our own empirical memristor model for validating emerging circuitry concepts and 2) new synergies with internationally leading groups (e.g. Waser's and Valov's groups in Aachen, Xu's group in NUDT, Andreou's group in John Hopkins, Indiveri's group in ETH and Legenstein's group in TU-Graz).
Sectors Aerospace, Defence and Marine,Education,Electronics,Energy,Healthcare,Security and Diplomacy

Description This work demonstrated for the first time that the physical mechanisms causing the inherent unreliability of Resistive Random Access Memory can be advantageously employed to reliably mimic synaptic connections; the necessary components for establishing brain-inspired electronics. This work was funded resulted in Prof Chua, inventor of memristors, spending three extended visits within Professor Prodromakis's group (Distinguished RAEng, Leverhulme and Marie Currie Fellowships). Other significances include further funding from EU-FP7 (RAMP €2.1M, ICT-2013-10: 612058) and EPSRC (EP/K017829/1, £1.4M), > 15 invited talks (2012-2014), 1 Nature Materials Commentary and Professor Prodromakis's selection to chair CAS-FEST 2014.
First Year Of Impact 2013
Sector Education,Electronics,Energy,Healthcare
Impact Types Societal,Economic

Description Future and Emerging Technologies
Amount € 2,100,000 (EUR)
Funding ID 612058 
Organisation European Commission 
Department Seventh Framework Programme (FP7)
Sector Public
Country European Union (EU)
Start 11/2013 
End 05/2017
Description PhD Scholarships
Amount £75,000 (GBP)
Organisation AG Leventis Foundation 
Sector Charity/Non Profit
Country Cyprus
Start 06/2011 
End 06/2014
Description e-Futures XD
Amount £50,600 (GBP)
Funding ID EFXD12003 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Public
Country United Kingdom
Start 04/2012 
End 01/2013
Title Memristor model 
Description We have developed a novel nonlinear dopant drift memristor model that resolves the boundary issues existing in previously reported models that can be easily adjusted to match the dynamics of distinct memristive elements. With the aid of this model, we examined switching mechanisms, current-voltage characteristics, and the collective ion transport in two terminal memristive devices, providing new insights on memristive behavior. 
Type Of Material Computer model/algorithm 
Year Produced 2012 
Provided To Others? Yes  
Impact A number of groups have already employed our model in various studies. This has facilitated a wider exploitation on new unconventional circuits and applications that exploit memristive elements. 
Title Volatile memristor model 
Description We have developed a new memristor SPICE model that introduces volatile effects, which can render a rate-dependent bipolar nonvolatile switching operation. The model is demonstrated via a number of simulation cases and is benchmarked against measured results acquired by solid-state TiO2 ReRAM. 
Type Of Material Computer model/algorithm 
Year Produced 2014 
Provided To Others? Yes  
Impact Currently, no available SPICE memristor model accounts for both nonvolatile and volatile resistive switching characteristics, the coexistence of which has been recently demonstrated to manifest on practical ReRAM. Besides exploiting this model for studying RRAM retention, our model can for the first time facilitate a more biorealistic exploitation of RRAM cells as chemical synapse emulators; allowing to capture both short- and long-term dynamics that co-exist in real neural circuits. This opens up a new design paradigm for neuromorphic engineering. 
Description Ramp project 
Organisation University of Padova
Country Italy 
Sector Academic/University 
PI Contribution Provided nanoscale memristor prototypes.
Collaborator Contribution Provided biological data and neuromorphic circuits.
Impact Employed memristors as biological sensors (electron devices, circuit design and electrophysiology).
Start Year 2013
Description Ramp project 
Organisation University of Zurich
Country Switzerland 
Sector Academic/University 
PI Contribution Provided nanoscale memristor prototypes.
Collaborator Contribution Provided biological data and neuromorphic circuits.
Impact Employed memristors as biological sensors (electron devices, circuit design and electrophysiology).
Start Year 2013