Overcoming the Memory Bottleneck in Auxiliary Field Quantum Monte Carlo Simulations with Interpolative Separable...

Area of Science:

• Computational Chemistry
• Materials Science
• Quantum Physics

Background:

• Auxiliary field quantum Monte Carlo (AFQMC) simulations are powerful for studying electron correlation in materials.
• Large memory requirements pose a significant bottleneck for AFQMC, limiting its application to larger systems.
• Gaussian basis sets are commonly used in electronic structure calculations.

Purpose of the Study:

• To introduce and investigate interpolative separable density fitting (ISDF) as a technique to mitigate memory limitations in AFQMC.
• To assess the impact of ISDF on the memory scaling of AFQMC simulations.
• To validate the effectiveness of ISDF by applying it to a real material system.

Main Methods:

• Implementation of interpolative separable density fitting (ISDF) within the AFQMC framework.
• Analysis of memory scaling before and after ISDF implementation.
• Calculation of structural properties for carbon in the diamond phase using the developed method.

Main Results:

• ISDF successfully reduces the memory scaling of AFQMC simulations from O(N^4) to O(N^3), where N is a measure of system size.
• The developed ISDF-enhanced AFQMC method accurately reproduces the structural properties of carbon in the diamond phase.
• Comparison with existing computational methods and experimental data validates the approach.

Conclusions:

• ISDF is an effective strategy for overcoming memory bottlenecks in AFQMC.
• This advancement enables more efficient and scalable quantum Monte Carlo simulations for real materials.
• The method holds promise for future investigations of complex material properties.