Metadata-Version: 2.4
Name: elecext
Version: 1.3.35
Summary: MRCC CPU binding: concurrent runs (parall_n, parall_n_mpi, parall) get disjoint physical-core slots; inside a slot OMP_PROC_BIND=spread,close + OMP_MAX_ACTIVE_LEVELS=1 + OMP_WAIT_POLICY=active so the nested ccsd/(T) OpenMP teams use every core of the slot; OpenMP settings logged next to the CPU set
Requires-Python: >=3.11
Requires-Dist: numpy
Provides-Extra: scalable
Requires-Dist: psutil; extra == "scalable"
