TEMA: “Artificial Synapse: a Materials Science Approach”
EXPOSITOR: Dr. Pablo Levy, researcher, CNEA, CONICET, Universidad Nacional de San Martín, Argentina
ABSTRACT: En esta charla se demuestra la conexión entre resultados que se obtuvieron con la aplicación de pulsos eléctricos sobre pastillas cerámicas de LaPrCaMnO con electrodos de Plata a temperatura ambiente y los dispositivos electrónicos con "capacidades neuromórficas"; es decir, sistemas artificiales que tienen algún nivel de semejanza con la sinapsis entre neuronas biológicas. La sinapsis artificial que se presenta en microdispositivos de dióxido de titanio (TiO2) nos habilita a pensar en la implementación de sistemas con capacidad de aprendizaje.
EXPOSITOR: Dr. Ioannis Vourkas. Assistant Professor, Departamento de Electrónica, Universidad Técnica Federico Santa María, Chile
ABSTRACT: Resistive switching electronic devices (ReRAM devices or memristors) have been known ever since the 60s. However, owing to the physical realization of the Chua’s memristor by the Hewlett-Packard Laboratories in 2008, new research tracks and trends in modern circuit design have indeed been created. The memristor, a nanoscale, nonvolatile, two-terminal resistive device whose resistance changes depending on the input signal applied to its terminals, is currently being explored for several emerging applications regarding upgraded and novel, energy-efficient digital/analog implementations such as nonlinear (chaotic) circuits, storage systems, logic circuits, neuromorphic and generally unconventional circuit architectures. This talk covers a timely topic of academic and industrial interest, aiming to stimulate further research on memristive devices, circuits, and systems. It particularly considers the design and development of nanoelectronic circuits, systems and computing architectures focusing on memristor as the main storage and computing element.
EXPOSITOR: Victor Grimblatt, R&D Group Director and General Manager, Synopsys Chile R&D Center
“Innovation runs at a scarily fast pace.” [William H. Gates III].
Our industry is now moving into the 5/3 nanometer nodes, which will require an unprecedented level of innovation and collaboration: new devices and materials are emerging that may replace FinFET at 3 nanometers, to say nothing about the manufacturing equipment; at the same time, sheer complexity continues to increase, as it has been the norm with the progress of Moore’s Law: several consumer products, integrating tens of billions of transistors, and manufactured in hundreds of millions of units, have successfully hit the market delivering unparalleled power and performance marks.
After decades of domination by general purpose CPU and GPU, innovation is disrupting also computing architectures: massively parallel Tensor Processing Units (TPU) have demonstrated that a computer can learn from past experience and, for example, become a chess Grandmaster in less than four hours, or classify zillions of images with surprising accuracy and speed.
The computing and memory requirements of Artificial Intelligence (AI) applications greatly exceed the capabilities of current electronics and are unlikely to be met by isolated improvements in transistors, data storage technologies or integrated circuit architectures alone.
Josephson Junction-based superconducting electronics promises to innovate High-Performance Computing (HPC) by delivering 100X more performance using 100X less power and is the foundation for a new class of computers, based on the laws of quantum physics. A 72 qubits Quantum Computer (QC), and an initiative to make cloud-based QC commercially available for businesses and research have been recently announced; QC may change the landscape of finance, imaging diagnostics, meteorology, pharmacology and, of course, security all the way.
This talk offers an overview of the key disruptive innovations that are changing the landscape of our industry.”