Mixed Precision Fermi-Operator Expansion on Tensor Cores From a Machine Learning Perspective

We present a second-order recursive Fermi-operator expansion scheme using mixed precision floating point operations to perform electronic structure calculations using tensor core units. A performance of over 100 teraFLOPs is achieved for half-precision floating point operations on Nvidia's A100...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	arXiv.org 2021-01
Hauptverfasser:	Finkelstein, Joshua, Smith, Justin, Mniszewski, Susan M, Barros, Kipton, Negre, Christian F A, Rubensson, Emanuel H, Niklasson, Anders M N
Format:	Artikel
Sprache:	eng
Schlagworte:	Artificial neural networks Electron states Electronic structure Floating point arithmetic Machine learning Mathematical analysis Optimization Physics - Computational Physics Physics - Materials Science Physics - Quantum Physics Quantum mechanics Tensors Thermal expansion
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!