We are at the beginning of a data age. In 2016, 90% of the world’s data had been created within the previous two years. With the development of AI and data analytics, the growth of data continues to accelerate worldwide. The energy cost of storing and moving data is enormous, much higher than the cost of computing the data. As energy-efficient computing increasingly shifts from being compute-centric to being memory-centric, the corresponding demand for energy is growing at such a rapid pace that within a few years the power demand may overwhelm some developed nations. Non-volatile memory (NVM) technology could solve problems of huge energy consumption and could also remove the data transfer bottleneck, increasing speed significantly. The technology is also key for creating artificial synapses for neuromorphic computing.
A range of NVM technologies are being explored. Some are at early stages and include ferroelectric field-effect-transistors (FeFET), molecular memory, carbon, macromolecular, and Spin-orbit/Mott. Others are more advanced (some already in industry), e.g. Phase memory (PCM) for storage class memory (SCM), spin transfer torque RAM (STT-RAM) for embedded-NVM (e-NVM), and resistive RAM, RRAM for both SCM and e-NVM.
This special topic will focus on emerging memory types and the advanced new materials science engineering required to make them suitable for NVM and neuromorphic computing applications.
Topics covered include, but are not limited to:
- Resistive switching
- Non-volatile memory
- Phase change memory
- 2D memory
- Ferroelectric memory
- Neuromorphic computing
Stuart Parkin, Max Planck Institute for Microstructure Physics
Themis Prodromakis, University of Southampton
Chang-Beom Eom, University of Wisconsin-Madison
Jordi Sort, Autonomous University of Barcelona
Judith Driscoll, University of Cambridge
Bhagwati Prasad, Western Digital