Terms and Conditions
By using GANSU2, you agree to the following terms.
Disclaimer of Warranty
GANSU2 is provided “as is”, without warranty of any kind, express or implied, including but not limited to the warranties of merchantability, fitness for a particular purpose, and non-infringement. In no event shall the authors or copyright holders be liable for any claim, damages, or other liability arising from the use of the software.
License and Permitted Use
A license key obtained through the User Portal permits any use of GANSU2, including commercial use, subject to the period and Max Size limit shown on your portal dashboard.
Without a license key, GANSU2 still runs, but the workload size is limited (Max Size). When the limit is reached the run stops with a message explaining how to obtain a key; nothing is silently reduced — GANSU2 never lowers the basis set, method, or accuracy to fit a limit.
Account- and license-specific terms are set out in the User Portal Account Terms of Service.
Prohibited Actions
The following actions are prohibited without prior written permission from the copyright holder:
- Reverse engineering, disassembly, or decompilation of the software or its libraries.
- Redistribution of the software, in whole or in part, to third parties.
- Sublicensing or granting any rights in the software to third parties.
- Circumventing the license activation, machine-binding, verification, or rate-limit mechanisms.
Third-Party Components
GANSU2 is derived from GANSU, licensed under the BSD 3-Clause License, Copyright (c) 2025-2026 Hiroshima University and Fujitsu Limited.
GANSU2 also includes or links the following third-party components:
- Eigen — MPL 2.0.
- OpenBLAS / LAPACK — BSD 3-Clause License (statically embedded).
- xxHash — BSD 2-Clause License. Copyright Yann Collet.
- OpenSSL — Apache License 2.0 (statically linked).
- GCC runtime (
libgfortran,libquadmath) — GPL v3 with the GCC Runtime Library Exception. - NVIDIA CUDA runtime (cuBLAS, cuSOLVER, cuSPARSE, NCCL, CUDA Runtime) — obtained on your machine (typically via the
nvidia-*-cu12packages) and used under NVIDIA’s license terms.