A container is a portable unit of software that combines the application and all its dependencies into a single package that is agnostic to the underlying host OS. In a high-performance computing (HPC) environment, containers remove the need for building complex environments or maintaining environment modules, making it easy for researchers and systems administrators to deploy their HPC applications.
There are more than a hundred containers spanning HPC, deep learning and machine applications available in the NGC catalog, NVIDIA’s hub of GPU-optimized HPC and AI applications. The containers available in the catalog are tested for performance, reliability and scalability. They are also screened for Common Exposure and Vulnerabilities (CVEs) and malware to ensure that they are ready for deployment in a production environment.
Below are some highlights of some of the new as well as updated containers that can run on both x86 and ARM platforms and fully support Singularity runtime.
New containers added to the catalog:
- TinkerHP is an MPI based, massively parallel package dedicated to long polarizable molecular dynamics simulations and to polarizable QM/MM.
- TorchANI is a PyTorch-Based Deep Learning Implementation of the ANI Neural Network Potentials.
- LBPM or Lattice Boltzmann Porous Media is a software package for simulating flow through porous media.
- NGC Pre-Flight Check light-weight tool verifies that the container runtime is set up correctly for GPUs and InfiniBand.
You can also find updated versions of the some of the key HPC applications:
- Chroma 2021.04
- Gromacs 2021
- Julia v2.4.2
- LAMMPS 10Feb2021
- NVIDIA HPC-SDK 21.3
- Relion 3.1.2
- ParaView 5.9.0
Pull your application container from NGC catalog now.