Human awareness is a fungal capacity, emerging from the brain’s ability to treat many sensory inputs. Simulating this function in electronic systems, which is usually referred to as nervous computing, has the ability to create very smart machines capable of supporting a wide range of daily applications, from independent vehicles to smart navigation systems. However, achieving nervous computing in practice represents great challenges, especially in the areas of energy efficiency, reliability and security.

The doctoral network React Marie Skłodowska-Curie (MSCA) addresses the challenges mentioned above by developing the nerve shape platform that is self-realized in its nature in terms of energy consumption, safe operation and system reliability.

Are you excited about the opportunity to join the doctoral network in the React, and participate in influential research while providing an international and practical perspective for your career?

As part of the RECT initiative, 15 researchers will be trained in the early stage (ESRS) through a comprehensive multidisciplinary program that extends on material science, physics in devices, computer engineering engineering, initial devices, translator design, simulation and simulation, as well as cybersecurity, reliability, and system verification.

The goal of this doctorate project is to develop the memory -based memory account structure (CIM) to enable online energy -saving online learning on the edge. Traditional edge devices are limited to restrictions in energy and cumin and the display of the frequency of memory, which constitute great challenges for actual time learning when using traditional VON Neumann structures. By taking advantage of the earning cell memory, which provides a high density and low leakage, and merges the mixed signature account directly into the memory sects, the proposed approach greatly reduces data movement and energy consumption. To ensure reliable operation under the real world conditions, mechanisms for detecting faults and effective recovery will also be evaluated and recovered, addressing analog computing and memory of noise, different process and soft errors. The primary goal is to design the CIM system to perform the main learning processes, such as the proliferation of expenses, weight and weight updates, within the same memory, thus enhancing energy efficiency and mathematical productivity. This architecture will be improved for light and adaptive learning tasks that are usually faced in edge scenarios, such as the sensor’s merger.

React provides a uniquely organized training environment, combining academic excellence and industrial cooperation. ESRS will benefit from close counseling by researchers and industry experts, while developing basic skills in scientific writing, research ethics, time management, and entrepreneurship. With the conclusion of the React project, the participants will be well equipped to follow up influential jobs throughout the academic and industry, as the React program works as a strong basis for their success in the future. More information about the project here can be found here https://project-rect.eu/.