Manu Nair
Position:
Postdoc --
ended Apr 2021
Email:
mnair@ini.uzh.ch
Work phone:
Location:
55g79-d19
I am interested in designing systems that compute better. My current work involves design of dedicated hardware for learning algorithms.
In total, I have about 6+ years of experience in designing computational systems. With my skill-set, I can build systems at various levels of abstraction, from transistor-level analog and digital design to high-level OOPS-based software systems.
I am currently working on:
1. Designing CMOS circuits for sensing and programming memristive devices.
2. Learning algorithms suitable for neuromorphic and similar architectures.
Supervisor
Giacomo Indiveri
Spin-Off Companies
Synthara Technologies
Publications
2019
Nair, Manu V and Indiveri, Giacomo
An Ultra-Low Power Sigma-Delta Neuron Circuit
,
2019 IEEE International Symposium on Circuits and Systems (ISCAS)
, 2019
2018
Melika Payvand, Manu Nair, Lorenz K Muller, Giacomo Indiveri
A neuromorphic systems approach to in-memory computing with non-ideal memristive devices: From mitigation to exploitation
,
Faraday Discussions
, 2018
Stefano Brivio, Daniele Conti, Manu V Nair, Jacopo Frascaroli, Erika Covi, Carlo Ricciardi, Giacomo Indiveri, Sabina Spiga
Extended memory lifetime in spiking neural networks employing memristive synapses with nonlinear conductance dynamics
,
Nanotechnology
, 2018
2017
L.K. Muller, M.V. Nair, G. Indiveri
Randomized unregulated step descent for limited precision synaptic elements
,
Proceedings-IEEE International Symposium on Circuits and Systems
, 2017
M.V. Nair, L.K. Muller, G. Indiveri
A differential memristive synapse circuit for on-line learning in neuromorphic computing systems
,
Nano Futures
, 1:(3), 2017
2016
Nair, Manu V; Dudek, Piotr
Practical gradient-descent for memristive crossbars
,
2015 International Conference on Memristive Systems (MEMRISYS)
, 2016
2015
Nair, Manu V; Dudek, Piotr
Gradient-descent-based learning in memristive crossbar arrays
,
2015 International Joint Conference on Neural Networks (IJCNN)
, 2015
back
back