[Q-e-developers] New GPU-accelerated Quantum ESPRESSO available

[Filippo Spiga]

Hello everyone,

I am pleased to announce you a new version of GPU-accelerated Quantum ESPRESSO based on CUDA FORTRAN going public today!

This project is still in its very early stages, very dynamic. This is a **pre-production** version of QE-GPU (v0.1), not all packages and not all functionalities are fully supported. Please refer to README.md for additional information how to download, compile and run the code. You need PGI compiler (17.4 Community Edition), you need  Open MPI and a data-centre grade NVIDIA TESLA Kepler (K20, K40, K80) or NVIDIA TESLA Pascal (P100) compute GPUs. *No other cards are supported* NVIDIA TESLA P100 is strongly recommend for its memory capacity and performance.

The GitHub page is here: https://github.com/RSE-Cambridge/qe-gpu
You can download the code directly from here: https://github.com/RSE-Cambridge/qe-gpu/archive/v0.1.tar.gz

If you have specific requests or specific input cases to test please email directly at spiga.filippo+qegpu at gmail.com

Cheers

--
Filippo SPIGA ~ http://fspiga.github.io ~ skype: filippo.spiga