I am a PhD Student at Imperial College London under the supervision of Dr. Christos-Savvas Bouganis. I am part of both the Circuits and Systems (CAS) group, the Intelligent Digital Systems Lab (iDSL) at Imperial College London, as well as the International Centre for Spatial Computational Learning (SpatialML).
My research focus is on Low-Power Machine Learning acceleration. I am exploring techniques which explictly target power reduction for Machine Learning applications. I am interested in novel coding and compression techniques for this domain. I am also working on an efficient framework for Convolutional Neural Network acceleration on FPGA devices.
Projects
- fpgaconvnet-hls: HLS backend for an FPGA-based Streaming Architecture for CNN Acceleration.
- fpgaconvnet-model: Parser and high-level performance and resource models for the fpgaconvnet-hls backend.
- samo: Design Space Exploration (DSE) tool for mapping CNN models to FPGAs for Streaming Architecture backends.
- pommel: Tool for estimating power and energy consumption of the memory subsystem in CNN Accelerators.
- def: ML-specific low-power coding scheme which reduces transitions on high-capacitance busses.
Publications
| [1] | A. Montgomerie-Corcoran*, Z. Yu* and C. Bouganis. SAMO: Optimised Mapping of Convolutional Neural Networks to Streaming Architectures In 2022 32nd International Conference on Field Programmable Logic and Applications (FPL), Aug 2022. [preprint] |
| [2] | A. Montgomerie-Corcoran and C. Bouganis. POMMEL: Exploring Off-Chip Memory Energy & Power Consumption in Convolutional Neural Network Accelerators In 2021 24th Euromicro Conference on Digital System Design (DSD), Sep 2021. [pdf, slides] |
| [3] | A. Montgomerie-Corcoran and C. Bouganis. DEF: Differential Encoding of Featuremaps for Low Power Convolutional Neural Network Accelerators In 2021 26th Asia and South Pacific Design Automation Conference (ASP-DAC), Jan 2021. [link, pdf, slides] |
| [4] | A. Montgomerie-Corcoran, S. I. Venieris, and C. Bouganis. Power-aware FPGA mapping of convolutional neural networks. In 2019 International Conference on Field-Programmable Technology (ICFPT), Dec 2019. [link, pdf, poster] |
Other Publications & Presentations
- A. Montgomerie-Corcoran. " FPGAConvNet & SAMO: Model-Specific Optimization of CNN Accelerators onto FPGAs" In 2022 6th Workshop on Reconfigurable Computing for Machine Learning (RCML), Sep 2022
- A. Montgomerie-Corcoran. " FPGAConvNet & SAMO: Model-Specific Optimization of CNN Accelerators onto FPGAs" In 2022 International Workshop on Research Open Automatic Design for Neural Networks (DAC-ROAD4NN), Jul 2022
- A. Montgomerie-Corcoran. "SAMO: Optimised Mapping of Convolutional Neural Networks to Streaming Architectures" In 2022 4th Workshop on Accelerated Machine Learning (AccML), Jun 2022
- A. Montgomerie-Corcoran. "fpgaConvNet: An Open-Source Streaming Architecture for Convolutional Neural Network Acceleration" In 2022 The 30th IEEE International Symposium On Field-Programmable Custom Computing Machines (FCCM), May 2022
- A. Montgomerie-Corcoran, D. Vink and A. Rajagopal. "The Hidden Environmental Cost of Machine Learning" In 2021 IEEE Conference on Advances in Communications, Devices and Systems (ACDS), Sep 2021