Hello, I'm Navdeep Singh Dhindsa

Research Interests

My primary research centers on investigating strong interactions in particle physics, with a key emphasis on employing lattice gauge theory as a fundamental methodology. As a newly appointed Postdoc, my current engagement involves numerical simulations of quantum chromodynamics (QCD) through lattice techniques. Specifically, my focus involves conducting lattice QCD studies on hadrons composed predominantly of heavy quarks. For these studies I am learning to implement Chroma package which adds on to my experience of working with MILC codes.

Throughout my doctoral research, I extensively explored the lattice framework for the non-perturbative study of field theories through numerical computations. Through engaging in this work, I have cultivated extensive proficiency in a diverse range of lattice techniques that hold broad applicability. Exploring field theories on a spacetime lattice is particularly intriguing, offering insights into coupling regimes beyond the reach of perturbation theory. My initial focus was on investigating non-perturbative supersymmetry breaking in lower dimensions. This exploration evolved into an examination of supersymmetric gauge theories including matrix models, with an emphasis on understanding their holographic duals through gauge/gravity duality. My current interest on these lines involves the exploration of maximally supersymmetric theories on the lattice, particularly in 1D, with a focus on the BFSS model and its mass-deformed BMN model.

My expertise includes working with the MILC-based SUSY LATTICE code, which is based on the MPI framework. In the future, I plan to address its computational challenges by parallelizing matrix degrees of freedom along with lattice sites to simulate higher-dimensional theories effectively. I generally use `Bash' and `Python' as tools to analyse the data generated in my projects.

Another avenue of interest is tackling problems related to lattice field theories, such as the sign problem. While I have experience with Monte Carlo simulations, I am also familiar with other numerical tools like the complex Langevin method (CLM) and Tensor Networks (TN). But my preference is for enhancing Monte Carlo simulations in real-world dimensions, as generalizing CLM and TN in higher dimensions is another tedious task. However, I am still open to working with the other methods if there is an exciting problem that these methods can solve better.

Another issue while working with Monte Carlo or evolution methods in general is that large computational resources are required to get adequate results. Hence, I am also interested in other non-perturbative non-lattice methods which can solve these models as effeciently or better as evolution methods. Recently, I've started looking into the numerical bootstrap method for multi-matrix models. Generalizing this method for more matrices poses a significant challenge, but it offers quick convergence in the absence of analytic solutions. I am trying to use this method to solve various matrix models at temperatures/couplings which are not accessible to Monte Carlo due to various reasons. My another short-term goal involves understanding the gravity duals by treating field theories with unit gauge links on the lattice.

Beyond the listed interests, I am eager to explore emerging approaches in lattice field theory, including the application of Machine Learning to train fields and the potential of Quantum Computing for studying Hamiltonian time evolution without concerns about simulation challenges like the sign problem.