FREEMIND · Ferroelectric REsistors as Emerging Materials for Innovative Neuromorphic Devices
Horizon 2020 — Marie Skłodowska-Curie Actions
- Duration
- 2019-06-01 → 2021-05-31
- EU contribution
- €191,149
- Participants
- 1
- Scheme
- MSCA-IF-EF-SE
Lines connect the coordinator with its partners.
Results in brief
Ferroelectric REsistors as Emerging Materials for Innovative Neuromorphic Devices
Worldwide, data traffic is exploding. Internet of Things, commercial transactions, media… This is possible thanks to always more performant computers. However, the well-known “Moore’s law” governing this trend is now reaching a physical limit. It is now urgent to find new computing architectures. An efficient way to handle large datasets is to use Artificial Neural Networks algorithms, but the training of these networks on conventional computers is time and energy demanding. The overall objective of the research project “FREEMIND” is to develop a hardware dedicated to Artificial Neural Networks algorithms, functioning not in the digital but in the analog domain. More specifically, matrix-vector multiplications will be implemented in the shape of a cross-bar array of "memristors". Memristors are electrical components whose resistance can be changed in a non-volatile way by an stimulus, such as an electric field. Such components represent the "synapses" of the neural networks, in an analogy with the biologic synapses connecting neurons together in the brain. In "FREEMIND", the objective is to use a ferroelectric material to create a memristor. Highly strategic for IBM Research, the focus is on a CMOS compatible and scalable process.
Data: CORDIS, © European Union
Project objective
Neuromorphic computing such as deep-learning algorithms arise as a promising solution to treat the exploding amount of data generated worldwide, but at the cost of expensive time and energy budget on conventional hardware. There is an urgent need for a low-power, compact neuromorphic chip that can support bio-inspired computing: a network of cells collocating storage (non-volatility) and computing (synaptic plasticity). This work proposes to achieve such cell using a ferroelectric resistor, down-scaled to nanometer thickness to allow direct electron tunneling through the ferroelectric barrier (ferroelectric tunnel junction): the learning functionality (i.e. synaptic plasticity) will be implemented through the control of the distribution of the (non-volatile) ferroelectric domains. The fellow will bring her expertise in ferroelectric tunnel junctions and will combine it with IBM’s expertise in device and circuits integration and characterization, making use of the state-of-the-art equipment offered by their research center. In order to accelerate the creation of an end-to-end neuromorphic device, she will lead a collaboration with ETH Zurich and will benefit from their expertise in predictive physics-based modeling. The research project will aim at: (i) the demonstration of a non-volatile and plastic ferroelectric “synapse” made of Hf0.5Zr0.5O2 – a fully CMOS-compatible material, (ii) the development of models of individual cells and of a neural network and (iii) providing design guidelines for neuromorphic hardware based on this technology. The outcome will be to evaluate performances not only of individual synapses but of a neural network as a whole. Through the Action, the fellow will not only aim at creating a novel technology; but also at leading an interdisciplinary research project uniting complementary actors for an innovative solution to a society challenge.
Original text from CORDIS.
Participants
- IBM RESEARCH GMBH · RUESCHLIKONCoordinatorSwitzerland
Links
Data: CORDIS, © European Union