Employment at TIMA

Internship proposals


Behavioral-Level Fault Modelling for Spiking Neural Networks

Host: TIMA Laboratory - AMfoRS team - 46 avenue Félix Viallet - 38031 GRENOBLE Cedex

Start Date: February 3rd, 2020

Duration: 6 months

Profile: The main objective of this internship is to analyze the possible SNN faults and provide a dictionary of fault models and a classification of faults by their severity.

The computing performance needed by emerging electronic applications (such as Internet-of-Things and Big Data analytics) is posing a serious challenge to current computer architectures and technologies, which are required to provide increasing computing power while withstanding severe constraints on size, energy consumption and reliability. Conventional Von-Neumann architectures and memories are not likely to fulfil all the needs of modern applications, due to inherent technological and conceptual limitations. Hence, in order to be at the forefront of the electronic industry in terms of design and manufacturing capabilities, it is essential to focus research and innovation efforts on the development of novel non-Von Neumann architectures enabled by emerging technology devices. In this context, the neuromorphic computing paradigm has a huge potential when it makes use of emerging NV technologies (STT-MRAM, memristors), however, reliable and testable HW designs enabling the neu romorphic computing are still missing.
This internship is concerned with the following research areas: (i) emerging memory technologies (memristors and/or spintronic devices) used in a non-Von Neumann context, (ii) hardware dependability (robustness, reliability and test) and design-for-dependability, (iii) hardware implementations of bio-inspired neural networks (Spiking Neural Networks).

See complete information

Contact person: Elena Ioana VATAJELU (ioana.vatajelu@univ-grenoble-alpes.fr)

Remuneration: 512 € / month

Level: Master 2 / Dernière année d'école d'ingénieur

 

 

Design and Evaluation of a Hardware-implemented Spiking Neural Network (SNN) with Spike-Timing-Dependent-Plasticity (STDP) learning

Host: TIMA Laboratory - AMfoRS team - 46 avenue Félix Viallet - 38031 GRENOBLE Cedex

Start Date: February 3rd, 2020

Duration: 6 months

Profile: The main objective of this internship is to design and evaluate a HW-implemented SNN using analog leaky-integrate and fire neurons and spintronic synapses.

Conventional Von-Neumann architectures and memories are not likely to fulfil all the needs of modern applications, due to inherent technological and conceptual limitations. Hence, in order to be at the forefront of the electronic industry in terms of design and manufacturing capabilities, it is essential to focus research and innovation efforts on the development of novel non-Von Neumann architectures enabled by emerging technology devices. In this context, the neuromorphic computing paradigm has a huge potential when it makes use of spintronic NV technologies.
This internship is concerned with the following research areas: (i) emerging memory technologies (memristors and/or spintronic devices) used in a non-Von Neumann context, (ii) hardware implementations of bio-inspired neural networks (Spiking Neural Networks).

See complete information

Contact person: Elena Ioana VATAJELU (ioana.vatajelu@univ-grenoble-alpes.fr)

Remuneration: 512 € / month

Level: Master 2 / Dernière année d'école d'ingénieur