The AMBER software for biomolecular simulations has a long-standing history of QM/MM implementations that couple the strength of quantum chemistry with the speed of molecular mechanics and thus enable simulations of local molecular properties and reaction mechanisms in complex environments such as enzymes and other condensed phase systems. Supported Hamiltonians range from built-in efficient but approximate semiempirical and density functional tight-binding methods to accurate but computationally more expensive <i>ab initio </i>methods via robust interfaces to external quantum chemistry software packages. Here we present an integration of AMBER with the open source Quantum Interaction Computational Kernel (QUICK) program via an application programming interface (API) that can also be used for integration with other molecular simulation software. We present recent enhancements to QUICK, including multi-GPU enabled Hartree–Fock, long-range electrostatics and density functional theory (DFT) energies, GPU optimized one- and two electron integrals and nuclear gradients with Gaussian basis sets. This leads to a performance on a single data center GPU that exceeds the performance of the corresponding MPI code on a state-of-the-art multi-core CPU compute node. Parallel scaling across multiple GPUs is excellent, thus enabling efficient large-scale QM/MM simulations.

Current status of AMBER coupled with QUICK

Where thousands of chemistry professionals meet to share ideas and advance scientific and technical knowledge. Spring 2023 abstracts, posters, and presentations will launch at the end of March.

ACS Spring 2023

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.