Memristive Devices Toward Smart Technical Systems
Termin:
30.09.2023
Fördergeber:
Deutsche Forschungsgemeinschaft (DFG)
A non-volatile memory resistor (short: memristor) is a revolutionary type of electronic memory device that combines the best features of conventional memory technologies. It has two terminals, and its resistance can be programmed and retained even after the device is powered off, making it a non-volatile memory. ReRAMs, Ferro-Electric Memories, Phase Change Memories, Atomic Switches etc. are representative examples of memristor device technologies that currently are actively researched due to their non-volatility, high density, high switching speed and low power consumption. At the same time, volatile memristors offer new possibilities for emulating brain dynamics and therefore building compact low-power bio-inspired systems. Based on these features, memristors are currently considered for a variety of applications, including memory, logic, sensing, in-memory computing, neuromorphic computing and machine learning, making them a versatile choice for AI, IoT devices and Cyber-Physical Systems. Additionally, their unique electrical properties and compatibility with complementary metal-oxide-semiconductor (CMOS) technology make them a promising candidate for integration into existing electronic systems. However, their commercialisation and widespread use in electronics is still in its early stages, and more research and development are needed to fully realise their potential. The aim of this Priority Programme is to demonstrate the increasingly important role of memristors, to enable the derivation of new results in basic research, to verify the performance of memristive systems in applications and to further develop new memristive technology in shaping the future of electronics.
The development of memristor technology requires expertise from multiple disciplines. The MemrisTec programme promotes the interdisciplinary collaboration among researchers from the fields of Materials Science, Device Physics, Device and Circuit modelling, Circuit Design, Computer Architecture, Neuromorphic Computing, Artificial Intelligence and Analog Signal Processing. The goal is to advance the state of the art in memristor technology by overcoming technical challenges and limitations. Furthermore, this multidisciplinary approach will enable a top-to-bottom and/or bottom-to-top approach towards a common application-oriented memristor-based system that is in tune with the demands of the market. Investigating the underlying physical properties of memristors and their behaviour under various conditions will enable the identification of the most appropriate memristor materials with the desired electrical and memory properties. New memristor devices will be developed that fulfil the performance and reliability requirements for storing, recalling and processing data in novel designs of computing systems that mimic the behaviour of the human brain.
Importantly, since a strong theoretical foundation is essential for the successful design of memristor-based circuits and systems and for the development of new applications that leverage the unique properties of memristors, each MemrisTec project proposal must include a partner engaged in theoretical investigations. A solid understanding of the fundamental principles and behaviour of memristors is essential for circuit design, device selection and system architecture.
The Priority Programme MemrisTec focuses on the following topics:
- Design of innovative adaptive analogue circuits
- Development of in-memory computing arrays
- Realisation of sensing devices
- Memristive (neuromorphic) circuits
- Hybrid approaches for in-memory computing
Project proposals should envisage the creation of memristor-centred, bio-inspired, reconfigurable, energy-efficient and real-time computing systems. It must be noted that memristor device technologies based on magnetic effects will not be considered, since they are the object of extensive research elsewhere. The same applies to common non-memristive approaches to neuromorphic system design, non-full custom ASICs, fully digital electronics and von Neumann hardware architectures, all of which are excluded from this Priority Programme.
Referring to the German-American DFG roundtable discussion on memristive systems in Dresden in February 2023, it is noted that project parts carried out in other countries can be funded by the DFG if they promise to add value to the projects being pursued under the Priority Programme in Germany and if they make an important additional contribution to achieving the objectives of the Priority Programme. Note that proposals that include principal investigators from abroad can only be accepted if the scientific coordinator of the Priority Programme has explicitly acknowledged the aforementioned requirements at the time of proposal submission.
Proposals must be written in English and submitted to the DFG by 12 July 2023. Please note that proposals can only be submitted via elan, the DFG’s electronic proposal processing system.
Further Information:
https://www.dfg.de/foerderung/info_wissenschaft/info_wissenschaft_23_33/
The development of memristor technology requires expertise from multiple disciplines. The MemrisTec programme promotes the interdisciplinary collaboration among researchers from the fields of Materials Science, Device Physics, Device and Circuit modelling, Circuit Design, Computer Architecture, Neuromorphic Computing, Artificial Intelligence and Analog Signal Processing. The goal is to advance the state of the art in memristor technology by overcoming technical challenges and limitations. Furthermore, this multidisciplinary approach will enable a top-to-bottom and/or bottom-to-top approach towards a common application-oriented memristor-based system that is in tune with the demands of the market. Investigating the underlying physical properties of memristors and their behaviour under various conditions will enable the identification of the most appropriate memristor materials with the desired electrical and memory properties. New memristor devices will be developed that fulfil the performance and reliability requirements for storing, recalling and processing data in novel designs of computing systems that mimic the behaviour of the human brain.
Importantly, since a strong theoretical foundation is essential for the successful design of memristor-based circuits and systems and for the development of new applications that leverage the unique properties of memristors, each MemrisTec project proposal must include a partner engaged in theoretical investigations. A solid understanding of the fundamental principles and behaviour of memristors is essential for circuit design, device selection and system architecture.
The Priority Programme MemrisTec focuses on the following topics:
- Design of innovative adaptive analogue circuits
- Development of in-memory computing arrays
- Realisation of sensing devices
- Memristive (neuromorphic) circuits
- Hybrid approaches for in-memory computing
Project proposals should envisage the creation of memristor-centred, bio-inspired, reconfigurable, energy-efficient and real-time computing systems. It must be noted that memristor device technologies based on magnetic effects will not be considered, since they are the object of extensive research elsewhere. The same applies to common non-memristive approaches to neuromorphic system design, non-full custom ASICs, fully digital electronics and von Neumann hardware architectures, all of which are excluded from this Priority Programme.
Referring to the German-American DFG roundtable discussion on memristive systems in Dresden in February 2023, it is noted that project parts carried out in other countries can be funded by the DFG if they promise to add value to the projects being pursued under the Priority Programme in Germany and if they make an important additional contribution to achieving the objectives of the Priority Programme. Note that proposals that include principal investigators from abroad can only be accepted if the scientific coordinator of the Priority Programme has explicitly acknowledged the aforementioned requirements at the time of proposal submission.
Proposals must be written in English and submitted to the DFG by 12 July 2023. Please note that proposals can only be submitted via elan, the DFG’s electronic proposal processing system.
Further Information:
https://www.dfg.de/foerderung/info_wissenschaft/info_wissenschaft_23_33/