Simulation and implementation of novel deep learning hardware architectures for resource constrained devices

Lammie, Corey (2022) Simulation and implementation of novel deep learning hardware architectures for resource constrained devices. PhD thesis, James Cook University.

[img]
Preview
PDF (Thesis)
Download (6MB) | Preview
View at Publisher Website: https://doi.org/10.25903/4ed1-gs60
 
277


Abstract

Corey Lammie designed mixed signal memristive-complementary metal–oxide–semiconductor (CMOS) and field programmable gate arrays (FPGA) hardware architectures, which were used to reduce the power and resource requirements of Deep Learning (DL) systems; both during inference and training. Disruptive design methodologies, such as those explored in this thesis, can be used to facilitate the design of next-generation DL systems.

Item ID: 77815
Item Type: Thesis (PhD)
Keywords: Circuit simulation, CNN, Computer architecture, Convolution, Convolutional neural networks, Deep Learning, Device modeling, EEG, Electroencephalography, Feature extraction, Hardware, In-Memory Computing, Memristive Crossbar Array, Memristors, Optimization, Performance evaluation, Prediction algorithms, PyTorch, ReRAM, RRAM, Seizure Detection, Seizure Prediction, Stochastic processes, Switches, Training
Related URLs:
Copyright Information: Copyright © 2022 Corey Lammie.
Additional Information:

Seven publications arising from this thesis are stored in ResearchOnline@JCU, at the time of processing. Please see the Related URLs. The publications are:

[Chapter 2] Rahimi Azghadi, Mostafa, Lammie, Corey, Eshraghian, Jason K., Payvand, Melika, Donati, Elisa, Linares-Barranco, Bernabe, and Indiveri, Giacomo (2020) Hardware implementation of deep network accelerators towards healthcare and biomedical applications. IEEE Transactions on Biomedical Circuits and Systems, 14 (6). pp. 1138-1159.

[Chapter 3] Lammie, Corey, Xiang, Wei, and Rahimiazghadi, Mostafa (2022) Modeling and simulating in-memory memristive deep learning systems: an overview of current efforts. Array, 13. 100116.

[Chapter 4] Li, Chenqi, Lammie, Corey, Dong, Xuening, Amirsoleimani, Amirali, Azghadi, Mostafa Rahimi, and Genov, Roman (2022) Seizure Detection and Prediction by Parallel Memristive Convolutional Neural Networks. IEEE Transactions on Biomedical Circuits and Systems, 16 (4). pp. 609-625.

[Chapter 5] Lammie, Corey, Eshraghian, Jason K., Lu, Wei D., and Rahimi Azghadi, Mostafa (2021) Memristive stochastic computing for deep learning parameter optimization. IEEE Transactions on Circuits and Systems II: Express Briefs, 68 (5). pp. 1650-1654.

[Chapter 6] Lammie, Corey, Xiang, Wei, Linares-Barranco, Bernabé, and Rahimi Azghadi, Mostafa (2022) MemTorch: an open-source simulation framework for memristive deep learning systems. Neurocomputing, 485. pp. 124-133.

[Chapter 7] Lammie, Corey, Rahimiazghadi, Mostafa, and Ielmini, Daniele (2021) Empirical metal-oxide RRAM device endurance and retention model for deep learning simulations. Semiconductor Science and Technology, 36 (6). 065003.

[Chapter 8] Lammie, Corey, Olsen, Alex, Carrick, Tony, and Rahimi Azghadi, Mostafa (2019) Low-power and high-speed deep FPGA inference engines for weed classification at the edge. IEEE Access, 7. pp. 51171-51184.

Date Deposited: 28 Feb 2023 04:15
FoR Codes: 40 ENGINEERING > 4007 Control engineering, mechatronics and robotics > 400706 Field robotics @ 30%
46 INFORMATION AND COMPUTING SCIENCES > 4611 Machine learning > 461103 Deep learning @ 35%
40 ENGINEERING > 4018 Nanotechnology > 401804 Nanoelectronics @ 35%
SEO Codes: 28 EXPANDING KNOWLEDGE > 2801 Expanding knowledge > 280110 Expanding knowledge in engineering @ 35%
28 EXPANDING KNOWLEDGE > 2801 Expanding knowledge > 280115 Expanding knowledge in the information and computing sciences @ 35%
24 MANUFACTURING > 2404 Computer, electronic and communication equipment > 240499 Computer, electronic and communication equipment not elsewhere classified @ 30%
Downloads: Total: 277
Last 12 Months: 60
More Statistics

Actions (Repository Staff Only)

Item Control Page Item Control Page