QChem-Trainer scales transformer-based neural network quantum state training to 1,536 Fugaku nodes with up to 8.41x speedup and 95.8% weak-scaling efficiency while matching FCI energies.
Wave function optimization in the variational monte carlo method,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.DC 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Large-scale Neural Network Quantum States for ab initio Quantum Chemistry Simulations on Fugaku
QChem-Trainer scales transformer-based neural network quantum state training to 1,536 Fugaku nodes with up to 8.41x speedup and 95.8% weak-scaling efficiency while matching FCI energies.