Presentation
Scalable Pitch-Constrained Neural Processing Unit for 3D Integration with Event-Based Imagers
TimeTuesday, December 7th3:30pm - 3:50pm PST
Location3018
SessionFrom Brains to Bits
Event Type
Research Manuscript
Presented In-Person
Emerging Models of Computation
Design
DescriptionEvent-based imagers are bio-inspired sensors presenting intrinsic High Dynamic Range and High Acquisition Speed properties.
However, noisy pixels and asynchronous readout result in poor energy-efficiency and excessively large output data rates.
In this work, we use Convolutional Spiking Neural Network filters to compensate these drawbacks and reduce output bandwidth by 10x.
We designed a neuromorphic core as a distributable block that benefits from 3D integration technology with direct and parallel access to 32x32 pixels, enabling reduced frequency operation.
Post-layout simulations depict a peak energy efficiency with 2.83pJ per Synaptic Operation (equivalent to 0.093fJ/event/pix) at the nominal literature input event rate.
However, noisy pixels and asynchronous readout result in poor energy-efficiency and excessively large output data rates.
In this work, we use Convolutional Spiking Neural Network filters to compensate these drawbacks and reduce output bandwidth by 10x.
We designed a neuromorphic core as a distributable block that benefits from 3D integration technology with direct and parallel access to 32x32 pixels, enabling reduced frequency operation.
Post-layout simulations depict a peak energy efficiency with 2.83pJ per Synaptic Operation (equivalent to 0.093fJ/event/pix) at the nominal literature input event rate.