DiHydrogen

Software · Wed Dec 16 00:00:00 EST 2020

DOI:https://doi.org/10.11578/dc.20200106.1· OSTI ID:1581178 · Code ID:33387

Maruyama, Naoya ^[1]; Essen, Brian Van ^[1]; Dryden, Nikoli J ^[1]; Benson, Thomas R ^[1]; Moon, Timothy Y ^[1]; Oyama, Yosuke ^[1]

---

1. Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)

DiHydrogen is the second version of the Hydrogen fork of the well-known distributed linear algebra library, Elemental. DiHydrogen is a GPU-accelerated distributed multilinear algebra interface with a particular emphasis on the needs of the scalable distributed deep learning training and inference. DiHydrogen is part of the Livermore Big Artificial Neural Network (LBANN) software stack.

• Short Name / Acronym: DiHydrogen
• Project Type: Open Source, Publicly Available Repository
• Site Accession Number: 1003667
• Software Type: Scientific
• Version: 0.1
• License(s): Apache License 2.0
• Research Organization: Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)
• Sponsoring Organization: USDOE National Nuclear Security Administration (NNSA)

  ---

  Primary Award/Contract Number:

  AC52-07NA27344
• DOE Contract Number: AC52-07NA27344
• Code ID: 33387
• OSTI ID: 1581178
• Country of Origin: United States