Mid-Range Scientific Computing
Computing is the backbone that supports biological and environmental research at the Environmental Molecular Sciences Laboratory (EMSL).
Through the EMSL User Program, researchers can submit proposals to access and utilize EMSL’s mid-range scientific computing and visualization resources and expertise. EMSL computing supports a broad range of scientific computing needs and resources that have been designed with user needs in mind.
Research Application
The Systems Modeling Integrated Research Platform supports EMSL’s flagship computing resources, which are vital to computational modeling and simulation and the utilization of machine learning (ML) and artificial intelligence (AI) for data-driven modeling.
Our scientific computing expertise intersects with all the other integrated research platforms to enable the Department of Energy’s Biological and Environmental Research program and EMSL Model–Experiment (ModEx) cycles.
Examples of research applications include:
- Scientific ML to build data-driven models from complex, heterogeneous data from EMSL instruments that unravel our fundamental understanding of biological and environmental processes and their coupling.
- Quantum chemistry and molecular dynamics simulations of molecules, surface interfaces, nanoparticles, and biological systems.
- Models and approaches that provide insights into the structure and function of biological systems, biogeochemical processes, and their regulation across scales, ranging from molecular to mesoscale.
- Methods to bridge EMSL imaging and time-series data and models to enable scientific discovery.
Available Resources
- Tahoma
- 98 terabytes of memory
- 160 nodes with 36 Intel© Xeon™ Gold 6254 cores and 384 GB/node memory
- 24 ML/AI nodes with 2 NVIDIA Tesla V100 32GB GPUs and 1536 GB/node memory
- Estimated peak performance of 1,015 teraflops
Tahoma supports the following software:
- GROMACS
- VASP
- CP2K
- NWChem
- LAMMPS
- PFLOTRAN
- AMBER
- MOLPRO
- SPAdes
- cisTEM
- RELION
- NAMD
- VMD
- WRF/WRF-Chem
- A broad range of Python and R-based workflows for data processing, visualization, and AI/ML calculations.
For ease of use, we offer a graphical user interface, Open OnDemand, that can be used by users to develop and test their codes and workflows, submit jobs, and visualize data.
EMSL staff will work with users to ensure all the derived software and workflows are version-controlled and containerized for portability and reproducibility.
A Tahoma user guide is also available for review and is continually updated.
Tips for Success
To best use these computing capabilities, it is recommended that you speak with the Systems Modeling Integrated Research Platform lead for specific guidelines applicable to your project. When reaching out, please include
- Objectives of the project
- Specifics about the software you want to use, including versions and alternatives
- Anticipated input and output files and formats.
We support a wide range of expertise. If you think your project may benefit from computing resources but aren’t sure where to start, talk to us.