Grant: Discovery Grants Program - Individual
University of Toronto
$83,000.00
Programmable photonics for compute acceleration
FY 2024-25
Summary
As artificial intelligence (AI) and machine learning (ML) permeates through society, the need for scalable, energy-efficient computing is urgent. Current ML hardware systems face computational bottlenecks due to limited communication and memory bandwidth. This is especially significant given the exponentially growing parameters in neural networks and the much slower pace of bandwidth improvement for chip communication. The power consumption for AI is escalating that the environmental footprint of training large machine learning models is dire. These developments necessitate a shift to new compute accelerator hardware that focuses on energy efficiency.
This proposal outlines an ambitious research program to develop integrated photonic technologies for sustainable and scalable computing systems. The research builds on a leading visible light silicon photonics platform established by the applicant. Central to the vision of an energy-efficient computing system is programmable photonics, which are photonic integrated circuits or microsystems that are reconfigurable on demand. Three research pillars are proposed: (1) Large-scale optical circuit switches for low-power photonic interconnects; (2) Optical in-memory computing architectures using commercial memory technologies for machine learning acceleration; (3) Integrated photonics for atomic qubit control in quantum processors.
The aims of this research are to complete proof-of-concept demonstrations within 5 years, and, in 10 years, to deliver integrated photonic solutions to overcome bottlenecks in communication, memory, and energy consumption in computing. Specific objectives include developing piezoelectric MEMS components, integrating lasers, and demonstrating optical switches with energy efficiencies of < 1 pJ/bit, optical in-memory computing at 15 fJ/FLOP, and integrated modulators for atomic control. The outcomes will position Canada at the forefront of photonics and stimulate economic growth. The research impact will be maximized through industry partnerships, international collaborations, and incorporating feedback from the broader photonics and computing communities. The proposed innovations have strong potentials to overcome major bottlenecks in today's computing systems to enable computing systems of the future.
Program
Discovery Grants Program - Individual
Recipient
Location
Ontario, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
RGPIN-2024-05029
Name nom
Poon, Joyce
Department département
Electrical and Computer Engineering
Organization id
31
Institution établissement
University of Toronto
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
83000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1510
Committee name en
Electrical and Computer Engineering
Committee name fr
Génie électrique et informatique
Area of application code
513
Area of application group en
Manufacturing processes and products
Area of application group fr
Procédés et produits industriels
Area of application en
Electrical and electronic machinery and equipment (including computer hardware)
Area of application fr
Machinerie et équipement électriques et électroniques (y compris matériel informatique)
Research subject code
2505
Research subject group en
Electrical and electronic engineering
Research subject group fr
Génie électrique et électronique
Research subject en
Photon devices
Research subject fr
Dispositifs photoniques
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Programmable photonics for compute acceleration
Keywords
integrated photonics; silicon photonics; waveguides; nanophotonics; photonic devices; quantum photonics; optical switching
Application summary
As artificial intelligence (AI) and machine learning (ML) permeates through society, the need for scalable, energy-efficient computing is urgent. Current ML hardware systems face computational bottlenecks due to limited communication and memory bandwidth. This is especially significant given the exponentially growing parameters in neural networks and the much slower pace of bandwidth improvement for chip communication. The power consumption for AI is escalating that the environmental footprint of training large machine learning models is dire. These developments necessitate a shift to new compute accelerator hardware that focuses on energy efficiency.
This proposal outlines an ambitious research program to develop integrated photonic technologies for sustainable and scalable computing systems. The research builds on a leading visible light silicon photonics platform established by the applicant. Central to the vision of an energy-efficient computing system is programmable photonics, which are photonic integrated circuits or microsystems that are reconfigurable on demand. Three research pillars are proposed: (1) Large-scale optical circuit switches for low-power photonic interconnects; (2) Optical in-memory computing architectures using commercial memory technologies for machine learning acceleration; (3) Integrated photonics for atomic qubit control in quantum processors.
The aims of this research are to complete proof-of-concept demonstrations within 5 years, and, in 10 years, to deliver integrated photonic solutions to overcome bottlenecks in communication, memory, and energy consumption in computing. Specific objectives include developing piezoelectric MEMS components, integrating lasers, and demonstrating optical switches with energy efficiencies of < 1 pJ/bit, optical in-memory computing at 15 fJ/FLOP, and integrated modulators for atomic control. The outcomes will position Canada at the forefront of photonics and stimulate economic growth. The research impact will be maximized through industry partnerships, international collaborations, and incorporating feedback from the broader photonics and computing communities. The proposed innovations have strong potentials to overcome major bottlenecks in today's computing systems to enable computing systems of the future.