Phaseless Auxiliary-Field Quantum Monte Carlo with Sparse Composite Tensors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for solving ground-state many-electron problems in quantum chemistry are computationally expensive and lack accuracy, with phaseless auxiliary-field quantum Monte Carlo (ph-AFQMC) facing challenges in scalability due to the rapid increase in the number of Slater determinants with the size of the active space.
Innovation Solution
The method involves obtaining multiple active spaces of a molecular system, determining coefficient tensors for each, combining them into a composite tensor using a tensor product, and using a cutoff value to reduce the number of Slater determinants in the trial wave function, thereby enhancing computational efficiency and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the size of the active space is increased to improve accuracy, then the number of Slater determinants increases rapidly, but the computational cost and complexity increase exponentially
Solution Approach 1:
The patent applies segmentation by dividing the molecular system into multiple active spaces, where each active space is treated separately to generate individual coefficient tensors. This segmentation prevents the exponential growth of computational complexity that would occur if all electrons and orbitals were treated in a single large active space, while still capturing the essential correlation effects in each region.
Solution Approach 2:
The patent merges the results from multiple segmented active spaces by computing a tensor product of the individual coefficient tensors to form a composite coefficient tensor. This combining approach allows the method to capture correlation effects across the entire system while maintaining the computational efficiency of treating smaller active spaces separately.
2Measurement precision
If the number of Slater determinants is increased to improve accuracy, then the representation of electron correlation improves, but the computational time and resources increase significantly
Solution Approach 1:
The patent applies partial action by using a cutoff value to truncate the composite coefficient tensor, keeping only the most significant elements above a certain threshold. This partial inclusion of Slater determinants maintains the essential accuracy for describing electron correlation while dramatically reducing the computational burden compared to including all possible determinants.
Solution Approach 2:
The patent changes the parameter of the coefficient tensor representation by applying a cutoff threshold to transform the full dense tensor into a sparse tensor. This parameter change allows the method to adapt the level of detail included based on computational resources available, maintaining accuracy where needed while reducing complexity elsewhere.
3Measurement precision
If a full coefficient tensor is used to maintain accuracy, then the representation of the wave function is complete, but the memory requirements and computational overhead become prohibitive
Solution Approach 1:
The patent extracts only the significant elements from the full coefficient tensor by applying a cutoff threshold. This extraction removes the less important elements that contribute minimally to the wave function accuracy, thereby reducing memory requirements and computational overhead while preserving the essential physics.
Solution Approach 2:
The patent changes the density parameter of the coefficient tensor from full/dense to sparse by applying a cutoff criterion. This parameter transformation maintains the accuracy of important correlations while reducing the quantity of data that must be stored and processed, making the method feasible for larger molecular systems.
Data Source
AI summary
Example embodiments of the present disclosure relate to a solution for ph-AFQMC with direct product multi-Slater determinants trial. Multiple active spaces of a molecular system may be obtained and multiple coefficient tensors may be determined respectively. A composite coefficient tensor may be determined based on a tensor product of the multiple coefficient tensors of the multiple active spaces, and a trial wave function may be further determined based on the composite coefficient tensor and a cutoff value. As such, the multiple coefficient tensors for the multiple active spaces may be determined, thus the computation can be reduced. Additionally, since a cutoff value is used, the composite coefficient tensor is a sparse tensor and the number of Slater determinants may be reduced. Further, the determined trial wave function may be further used in a ph-AFQMC algorithm, and a balance between accuracy and efficiency may be achieved.