The reliance on Si for wireless communication, computing and consumer applications and for automative and other industries is under threat. Si is the key material in digital hardware used for computing and sensing applications today. However, global demand for Si-based sensor technologies is experiencing an exponential growth with forecasts of ~45 trillion sensors in 2032 that will generate > 1 million zettabytes (1027 bytes) of data per year. To alleviate this demand, a cross-disciplinary and cross-functional approach in connection to achieving fundamental breakthroughs using analog hardware, as an alternative to Si-based conventional digital hardware, should be pursued.
One emerging application of this kind of analog hardware is neuromorphic computing or in-memory computing. Traditional computing, particularly when it comes to implementing machine-learning and neural-network algorithms for data-classification tasks, suffers from von Neumann bottleneck (memory and computing are physically separate in the traditional computer (CPU or GPU), leading to time and energy consumption for shuffling data between these two units). As a result, efforts are being made to implement neural networks on analog hardware where memory and computing are physically intertwined. Such hardware is often referred to as neuromorphic (brain-inspired), or in-memory-computing hardware, and makes use of different materials outside the conventional Si-based CMOS technology, such as memristive materials, phase change materials, spintronics based materials etc.
Other related applications of analog hardware, which have captured a lot of attention of late, are: oscillator-based computing and stochastic/ probabilistic computing. These computing strategies have been proposed and sometimes demonstrated as alternatives to conventional digital computing, and even quantum computing, for solving different optimization problems and data-classification problems. Oscillator-based computing and probabilistic computing also make use of emerging devices as auto-oscillators, random-number generators, etc..
This Research Topic seeks to contribute to the state of the art on advanced materials, novel nanoelectronic devices, new system designs, and manufacturing strategies that are needed and expected to lead to breakthroughs in analog hardware research. The contributions will hopefully lead to new analog technologies based on in-memory computing, neuromorphic (brain-inspired) computing, oscillator-based computing, probabilistic computing, and smarter machine interfaces that operate from gigahertz to the THz regime.
The Research Topic welcomes articles about novel materials and devices development, new circuit and architecture design, and device modeling, with respect to (but not limited to) the following areas:
- Emerging materials and devices for neuromorphic/ in-memory computing
- Design of novel circuits, systems, and algorithms using emerging devices for neuromorphic/ in-memory computing
- Oscillator-based computing and probabilistic/ stochastic computing using emerging materials and devices
The reliance on Si for wireless communication, computing and consumer applications and for automative and other industries is under threat. Si is the key material in digital hardware used for computing and sensing applications today. However, global demand for Si-based sensor technologies is experiencing an exponential growth with forecasts of ~45 trillion sensors in 2032 that will generate > 1 million zettabytes (1027 bytes) of data per year. To alleviate this demand, a cross-disciplinary and cross-functional approach in connection to achieving fundamental breakthroughs using analog hardware, as an alternative to Si-based conventional digital hardware, should be pursued.
One emerging application of this kind of analog hardware is neuromorphic computing or in-memory computing. Traditional computing, particularly when it comes to implementing machine-learning and neural-network algorithms for data-classification tasks, suffers from von Neumann bottleneck (memory and computing are physically separate in the traditional computer (CPU or GPU), leading to time and energy consumption for shuffling data between these two units). As a result, efforts are being made to implement neural networks on analog hardware where memory and computing are physically intertwined. Such hardware is often referred to as neuromorphic (brain-inspired), or in-memory-computing hardware, and makes use of different materials outside the conventional Si-based CMOS technology, such as memristive materials, phase change materials, spintronics based materials etc.
Other related applications of analog hardware, which have captured a lot of attention of late, are: oscillator-based computing and stochastic/ probabilistic computing. These computing strategies have been proposed and sometimes demonstrated as alternatives to conventional digital computing, and even quantum computing, for solving different optimization problems and data-classification problems. Oscillator-based computing and probabilistic computing also make use of emerging devices as auto-oscillators, random-number generators, etc..
This Research Topic seeks to contribute to the state of the art on advanced materials, novel nanoelectronic devices, new system designs, and manufacturing strategies that are needed and expected to lead to breakthroughs in analog hardware research. The contributions will hopefully lead to new analog technologies based on in-memory computing, neuromorphic (brain-inspired) computing, oscillator-based computing, probabilistic computing, and smarter machine interfaces that operate from gigahertz to the THz regime.
The Research Topic welcomes articles about novel materials and devices development, new circuit and architecture design, and device modeling, with respect to (but not limited to) the following areas:
- Emerging materials and devices for neuromorphic/ in-memory computing
- Design of novel circuits, systems, and algorithms using emerging devices for neuromorphic/ in-memory computing
- Oscillator-based computing and probabilistic/ stochastic computing using emerging materials and devices