Exploiting floating point emulation in tensor contractions in relativistic coupled cluster theory
Abstract
In this manuscript we outline a computational strategy to leverage a vendor-optimized implementation of the Ozaki scheme I for floating point emulation as part of an implementation of the Transpose-Transpose-GEMM-Transpose (TTGT) pattern for tensor contractions, and assess its potential to accelerate two of the most expensive tensor contractions appearing in relativistic coupled cluster theory in high-end accelerators such as the NVIDIA Blackwell B200 system in which there is a strong differential in the performance of computations for double and lower precision datatypes. Our results indicate that this emulation-based TTGT strategy can provide significant speedups compared to runs without emulation, and can be competitive with state-of-the-art, vendor-optimized tensor contraction libraries using native double precision as compute datatype. More importantly, our scheme provides a path towards easily allowing users to tune the desired accuracy in the tensor contraction steps, including cases where precision between single and double is sufficient for the application.
Related
- Precise Determination of Electric Dipole Polarizabilities of Sr Clock States Using Linear-Response and Equation-of-Motion Relativistic Coupled-Cluster Methods
- Exploiting Sn(ii)/Sn(iv) redox chemistry: a silyl-stannylene offering close-to-thermoneutral transfer hydrogenation
- Exploiting Carboxylate Effects to Enhance Cp*Co III -Catalyzed C–H Functionalization
- Systematic mapping of bimetallic catalyst combinations for floating catalyst chemical vapor deposition growth of single-walled carbon nanotubes
- Exploiting the B–N/C–C Isosterism Yields Nile‐Red Inspired Organoboron Dyes Capable of Singlet‐Oxygen Photosensitization
BibTeX
@misc{le2026exploiting,
title = {{Exploiting floating point emulation in tensor contractions in relativistic coupled cluster theory}},
author = {Maxime Le and Ryan Benazzouk and Jan Brandejs and Andre Severo Pereira Gomes},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.06533},
eprint = {2611.06533},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.06533},
note = {CC BY-SA 4.0},
}
RIS
TY - UNPB
TI - Exploiting floating point emulation in tensor contractions in relativistic coupled cluster theory
AU - Maxime Le
AU - Ryan Benazzouk
AU - Jan Brandejs
AU - Andre Severo Pereira Gomes
PY - 2026
DA - 2026/11/04
PB - arXiv
DO - 10.48550/arxiv.2611.06533
UR - https://arxiv.org/abs/2611.06533
AB - In this manuscript we outline a computational strategy to leverage a vendor-optimized implementation of the Ozaki scheme I for floating point emulation as part of an implementation of the Transpose-Transpose-GEMM-Transpose (TTGT) pattern for tensor contractions, and assess its potential to accelerate two of the most expensive tensor contractions appearing in relativistic coupled cluster theory in high-end accelerators such as the NVIDIA Blackwell B200 system in which there is a strong differential in the performance of computations for double and lower precision datatypes. Our results indicate that this emulation-based TTGT strategy can provide significant speedups compared to runs without emulation, and can be competitive with state-of-the-art, vendor-optimized tensor contraction libraries using native double precision as compute datatype. More importantly, our scheme provides a path towards easily allowing users to tune the desired accuracy in the tensor contraction steps, including cases where precision between single and double is sufficient for the application.
ER -