wxj6000
commented
1 month ago - Updated the benchmark data with qchem and psi4, including the script for generating benchmark data.
- Improved CPHF solver with dynamic slicing
- Improved logging system and dump GPU info and CUDA library info into log file.
- Fixed an issue when an array size exceeding int32
- Improved Krylov subspace solver with stable QR decomposition