Скачать с ютуб Samvaad Talk by Prof. Ananda Y R, Assistant Professor, IIIT-Bangalore (January 20, 2025) в хорошем качестве

Samvaad Talk by Prof. Ananda Y R, Assistant Professor, IIIT-Bangalore (January 20, 2025)

#samvaadtalks #IIITB #iiitbangalore Title: Advanced Memcomputing Devices Abstract: Memristor, memcapacitor, and meminductor are the three types of memory elements (memelements or memcomputing devices). The memristor is the fourth fundamental circuit element based on the missing relationship between two electrical quantities, the charge (q) and the flux (φ). The memristor is considered one of the most promising nano-devices among those currently being studied for possible use in future electronic systems. The best performance features include fast switching speed, high endurance and data retention, low power consumption, high integration density, and CMOS compatibility. Memristors are being explored as a potential technology to replace CMOS for logic-in-memory systems exploiting memristive nonvolatility. The logic-in-memory (LIM) operation is one of the prominent characteristic features of the memristor, which effectively solves the so-called memory wall problem in conventional von Neumann architecture. A memristive device is highly nonlinear and non-volatile, which makes this device is better storage element with greater data density than the existing memory devices. In addition, the memristor exhibits switching capability, which is more relevant for implementing logic gates, a realization of Boolean functions, and digital system design. Meminductors and memcapacitors are the two special classes of memelements that exhibit inductive and capacitive behaviour. The memcapacitor is the constitutive relation between the time integral of the charge (TIC) and the flux, and the meminductor is the time integral of the flux (TIF) and the charge. These memelements exhibit a pinched hysteresis loop (PHL), which indicates their nonvolatile behaviour capable of storing the data. Memcapacitors and meminductors are lossless devices and more power efficient than the memristor. The distinctive properties of memelements offer several advantages, such as high scalability and low power consumption, making them suitable for designing high-performance neuromorphic computing, programmable analog ICs, oscillators, filters, amplifiers, process analog information in artificial intelligence (AI) applications, adaptive learning circuits, spiking neural networks, chaotic oscillators, and several bio-inspired applications. These memelements are expected to be available soon due to the requirement of suitable materials and a new fabrication process. However, emulators are designed by research communities to realize memelements using available solid-state electronic devices when designing practical applications. The emulators presented in the literature require more active and passive elements, which increases the hardware complexity with limited operating frequency. It motivated us to propose area and power-efficient emulators capable of operating at high frequency to realize memelements using a minimum number of off-the-shelf circuit elements. Speaker's Bio: Ananda Y. R. completed his Ph.D. in 2023 from the Department of Electronics and Electrical Engineering, IIT Guwahati, Assam, India. He obtained a B.E. degree in Electronics and Communication Engineering and an MTech. degree in VLSI Design and Embedded Systems from Visvesvaraya Technological University, Belgaum, Karnataka, India, in 2011 and 2014, respectively. Before starting his Ph.D., he worked as an assistant professor at the Department of Electronics and Communication Engineering, Dayananda Sagar College of Engineering (DSCE), Bangalore from 2014 to 2017. Additionally, he served as a Junior Research Fellow (JRF) in a DST-SERB sponsored project titled "Performance and energy optimization in many-core processors using dynamic cooperation cache memory, NoC, and DRAM controller" at the Department of Computer Science and Engineering, IIT Guwahati, from 2017 to 2018. His research interests include Memcomputing Devices (Memelements), Analog and Mixed Signal Integrated Circuit Design, Semiconductor devices, Digital circuit design and FPGA implementation, and VLSI architectures. #IIITB #IIITBangalore #SamvaadTalks #IIITBTalks #SanvaadSeries #ResearchTalks