Sunny talks to Dr Charlotte Frenkel from the Technical University of Delft. Charlotte set records with a low-power neuromorphic chip she designed as part of her PhD. In this episode of Brains and Machines she talks to Dr Sunny Bains of University College London about what she’s learned about building simplicity into chips and integrity into benchmarks. After the interview, discussion follows with Giulia and Ralph.
Articles and Papers
- A 0.086-mm2 12.7-pJ/SOP 64k-synapse 256-neuron online-learning digital spiking neuromorphic processor in 28-nm CMOS
- Bottom-up and top-down approaches for the design of neuromorphic processing systems: tradeoffs and synergies between natural and artificial intelligence
- EvGNN: An event-driven graph neural network accelerator for edge vision
- ReckOn: A 28nm sub-mm2 task-agnostic spiking recurrent neural network processor enabling on-chip learning over second-long timescales
- Chameleon: A matmul-free temporal convolutional network accelerator for end-to-end few-shot and continual learning from sequential data
From the discussion
- An event-based digital compute-in-memory accelerator with flexible operand resolution and layer-wise weight/output stationarity
- The neurobench framework for benchmarking neuromorphic computing algorithms and systems
- Probabilistic synaptic weighting in a reconfigurable network of VLSI integrate-and-fire neurons
- An analog neural computer with modular architecture for real-time dynamic computations
- Working with neural networks
- Which model to use for cortical spiking neurons?
Episode Credits
Producer/Writer: Sunny Bains
Co-hosts: Sunny Bains and Giulia D’Angelo
Editor: Rose Gotto
Audio Production: Taylor Marvin, Coupe Studios Music and Sound Design
Music: 3 of Diamonds by Matt Harris
For EETimes.com
This episode is produced in conjunction with EETimes. Podcast reproduced with permission.
