Grant: Discovery Grants Program - Individual
McMaster University
$43,000.00
Smart Materials for Neuromorphic Computing
EF 2024-25
Résumé
For decades, the semiconductor industry has been driven by Moore's law, which states that the number of transistors on a microchip will double every 2 years. This law is facing a severe bottleneck where no further miniaturization of standard memory devices is possible. As such, a complete redesign of computer hardware technology using neuromorphic components is required to further computing technology. Such devices are not limited to binary logic, allowing them to emulate the human brain's learning behaviour. While contemporary algorithms can already perform such cognitive tasks, the energy for a single computational training activity remains orders of magnitude higher than the brain (13 kW compared to a brain's 20 W), resulting in a very high carbon footprint. Neuromorphic components can reduce this energy consumption to nearly that of the human brain. However, the materials they are composed of are not optimized, impeding them from mimicking bio-inspired signal pathways in larger artificial neuronal networks.
This research program's long-term objective is to develop smart materials for highly energy efficient and performant neuromorphic systems that can reproduce a human brain's cognitive capabilities. It focusses on the only two known ferroelectric materials suitable for large scale integration: Hf1-xZrxO2 and Al1-xScxN. Here, synaptic learning behaviour is directly emulated by an applied external voltage inducing resistance changes in the materials. By combining materials synthesis with advanced characterisation techniques, device fabrication, and their subsequent performance assessment, this research program's short-term objectives will be to (i) understand the interplay between material structure and microscopic characteristics, (ii) exploit the effect of external mechanical stress to further enhance ferroelectricity, (iii) fabricate ferroelectric tunnel junctions (FTJs) devices and understand their charge transport mechanisms, and (iv) assess the FTJs performances as artificial synapses and tune them to specific applications. Numerical simulations of the different physical effects will complement experimental results to identify the dominant mechanisms at play. To develop highly efficient artificial synapses, this program will determine how a material's microstructure dictates device performance, resulting in a recipe to tune the devices for marketable applications such as smart sensors, actuators, or energy harvesting devices.
As artificial intelligence becomes increasingly present in every Canadian's life, more powerful pattern recognition devices will have a significant social impact. In the current energy crisis, reducing the computational energy consumption required for machine learning algorithms will strongly benefit Canada's economy. This will in turn result in a lower carbon footprint, reducing climate change, and helping to protect Canada's environment.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
RGPIN-2023-04762
Name nom
Kolhatkar, Gitanjali
Department département
Engineering Physics
Organization id
27
Institution établissement
McMaster University
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2023
Award amount
43000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1511
Committee name en
Materials and Chemical Engineering
Committee name fr
Génie chimique et des matériaux
Area of application code
802
Area of application group en
Information and communication services
Area of application group fr
Services d'information et de télécommunications
Area of application en
Information systems and technology
Area of application fr
Systèmes d'information et technologie de l'information
Research subject code
2001
Research subject group en
Materials science and technology
Research subject group fr
Science et technologie des matériaux
Research subject en
Materials structure, properties and testing
Research subject fr
Structure, propriétés et essais des matériaux
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Smart Materials for Neuromorphic Computing
Keywords
neuromorphic devices; piezoelectric and ferroelectric materials; cmos-compatible; structural properties; nanoscale imaging; charge transport mechanisms; thin-film synthesis; electronic properties; artificial synapses; smart sensors
Application summary
For decades, the semiconductor industry has been driven by Moore's law, which states that the number of transistors on a microchip will double every 2 years. This law is facing a severe bottleneck where no further miniaturization of standard memory devices is possible. As such, a complete redesign of computer hardware technology using neuromorphic components is required to further computing technology. Such devices are not limited to binary logic, allowing them to emulate the human brain's learning behaviour. While contemporary algorithms can already perform such cognitive tasks, the energy for a single computational training activity remains orders of magnitude higher than the brain (13 kW compared to a brain's 20 W), resulting in a very high carbon footprint. Neuromorphic components can reduce this energy consumption to nearly that of the human brain. However, the materials they are composed of are not optimized, impeding them from mimicking bio-inspired signal pathways in larger artificial neuronal networks.
This research program's long-term objective is to develop smart materials for highly energy efficient and performant neuromorphic systems that can reproduce a human brain's cognitive capabilities. It focusses on the only two known ferroelectric materials suitable for large scale integration: Hf1-xZrxO2 and Al1-xScxN. Here, synaptic learning behaviour is directly emulated by an applied external voltage inducing resistance changes in the materials. By combining materials synthesis with advanced characterisation techniques, device fabrication, and their subsequent performance assessment, this research program's short-term objectives will be to (i) understand the interplay between material structure and microscopic characteristics, (ii) exploit the effect of external mechanical stress to further enhance ferroelectricity, (iii) fabricate ferroelectric tunnel junctions (FTJs) devices and understand their charge transport mechanisms, and (iv) assess the FTJs performances as artificial synapses and tune them to specific applications. Numerical simulations of the different physical effects will complement experimental results to identify the dominant mechanisms at play. To develop highly efficient artificial synapses, this program will determine how a material's microstructure dictates device performance, resulting in a recipe to tune the devices for marketable applications such as smart sensors, actuators, or energy harvesting devices.
As artificial intelligence becomes increasingly present in every Canadian's life, more powerful pattern recognition devices will have a significant social impact. In the current energy crisis, reducing the computational energy consumption required for machine learning algorithms will strongly benefit Canada's economy. This will in turn result in a lower carbon footprint, reducing climate change, and helping to protect Canada's environment.