Canada Spends Logo

Explore our Budget 2026 analysis

Grant: Discovery Grants Program - Individual

University of Victoria

$33,000.00

Sparse-rank factorization methods for quantum problems

FY 2023-24

Summary
Modeling how a material works is almost as good as actually making it. We need to eventually make materials to design new medicines, building materials, and computers that will propel future innovations. However, the cost to simulate a material before it is even made is so much less expensive that it can accelerate the time to make a new technology in a laboratory. In the case of materials, the correct theory to evaluate is quantum mechanics, the study of how very small objects behave and interact. As it turns out, the understanding of quantum mechanics is very good and very accurate. Predictions in quantum mechanics can match experiments out to 20 digits in some cases, an agreement that is not found in some other fields of science. Solving quantum problems is a difficulty. The problem has been proven to be so hard that it is not expected to ever find an algorithm for the full problem. Having said that, the time spent finding algorithms that work in limited cases or give approximate results are very worthwhile because they can produce answers that are close enough to experiments to give reliable predictions. Over the past 100 years, the understanding of how to solve quantum mechanics has improved a lot. This proposal is building on work that demonstrated that an algorithm derived in 1958 by the mathematician and physicist, Cornelius Lanczos, can be repurposed on a quantum computer to solve some problems that we have not been able to solve so far. That is to say, the classical computers that we use today are not fundamentally designed to properly implement this nearly 65 year old algorithm! However, on the quantum computer, the way that information is stored is different than on a classical computer. In order to represent Lanczos' algorithm on the classical computer, it would take a giant computer to represent. On the classical computer, the amount of storage space is exponentially compressed. In this proposal, I will extend work I have previously done on extending Lanczos' method to many situations that will be useful for designing materials and technologies with the quantum computer. More recently, in the last 30 years, a class of algorithms on the classical computer has also been implemented successfully and used to solve important problems. The new algorithm is called a tensor network and decomposes the quantum problem into a series of tensors that avoid the exponentially large size of representing the full problem. Introducing Lanczos' algorithms into tensor networks has the potential to reach farther and to solve quantum problems that other algorithms have difficulty solving at present. This proposal will allow me to extend my previous work on quantum algorithms to tensor networks and introduce a new class of algorithms based on Lanczos' work there. While we wait for the development of quantum computers to reach full maturity, these algorithms can provide advances in the near future for materials that modern society needs.
Program
Discovery Grants Program - Individual
Location
British Columbia, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data

As published

Application id
RGPIN-2023-05510
Name nom
Baker, Thomas
Department département
Department of Physics
Organization id
7
Institution établissement
University of Victoria
Province en
British Columbia
Province fr
Colombie-Britannique
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2023
Competition year année de concours
2023
Award amount
33000
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
1505
Committee name en
Physics
Committee name fr
Physique
Area of application code
1208
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Materials sciences
Area of application fr
Sciences des matériaux
Research subject code
3300
Research subject group en
Condensed matterphysics
Research subject group fr
Condensed matterphysics
Research subject en
Physique de la matière condensée
Research subject fr
Physique de la matière condensée
Application title
Sparse-rank factorization methods for quantum problems
Keyword
Density functional theory; Lanczos methods; Quantum algorithms; Quantum chemistry; Quantum computing; Quantum materials; Sparse-rank factorization methods; Tensor networks
Application summary
Modeling how a material works is almost as good as actually making it. We need to eventually make materials to design new medicines, building materials, and computers that will propel future innovations. However, the cost to simulate a material before it is even made is so much less expensive that it can accelerate the time to make a new technology in a laboratory. In the case of materials, the correct theory to evaluate is quantum mechanics, the study of how very small objects behave and interact. As it turns out, the understanding of quantum mechanics is very good and very accurate. Predictions in quantum mechanics can match experiments out to 20 digits in some cases, an agreement that is not found in some other fields of science. Solving quantum problems is a difficulty. The problem has been proven to be so hard that it is not expected to ever find an algorithm for the full problem. Having said that, the time spent finding algorithms that work in limited cases or give approximate results are very worthwhile because they can produce answers that are close enough to experiments to give reliable predictions. Over the past 100 years, the understanding of how to solve quantum mechanics has improved a lot. This proposal is building on work that demonstrated that an algorithm derived in 1958 by the mathematician and physicist, Cornelius Lanczos, can be repurposed on a quantum computer to solve some problems that we have not been able to solve so far. That is to say, the classical computers that we use today are not fundamentally designed to properly implement this nearly 65 year old algorithm! However, on the quantum computer, the way that information is stored is different than on a classical computer. In order to represent Lanczos' algorithm on the classical computer, it would take a giant computer to represent. On the classical computer, the amount of storage space is exponentially compressed. In this proposal, I will extend work I have previously done on extending Lanczos' method to many situations that will be useful for designing materials and technologies with the quantum computer. More recently, in the last 30 years, a class of algorithms on the classical computer has also been implemented successfully and used to solve important problems. The new algorithm is called a tensor network and decomposes the quantum problem into a series of tensors that avoid the exponentially large size of representing the full problem. Introducing Lanczos' algorithms into tensor networks has the potential to reach farther and to solve quantum problems that other algorithms have difficulty solving at present. This proposal will allow me to extend my previous work on quantum algorithms to tensor networks and introduce a new class of algorithms based on Lanczos' work there. While we wait for the development of quantum computers to reach full maturity, these algorithms can provide advances in the near future for materials that modern society needs.