Bridges-2 GPU

Bridges-2 GPU is the GPU-accelerated tier of Bridges-2, built from nodes that pair NVIDIA H100, L40S, or V100 GPUs (8 per node, 16 on the DGX-2) with two Intel Xeon CPUs and up to 2 TB of RAM. It is well suited to GPU-accelerated deep learning and molecular dynamics, and is often used for large-scale model training and inference and for biomolecular simulation. Its most heavily used software spans machine-learning frameworks and molecular-dynamics codes.

Get Support

GET AN ACCOUNT ON BRIDGES-2 GPU

Request an account

Submitting Jobs Documentation

You can run jobs at different sizes and durations on Bridges-2 GPU. The following lists the different queues that you can submit to, describing how many nodes you get, how long you can run, the type of resources you get, and the average wait time.

Jobs are submitted through Slurm. Unless you set otherwise, a job uses a one-hour walltime, up to a 48-hour maximum. Request GPUs by type and count with --gpus=type:n or --gres=gpu:type:n, where the type is one of h100-80, l40s-48, v100-32, or v100-16. In batch mode n is the total GPUs for the job (a multiple of 8); in an interactive session (interact) it is the GPUs per node. Pick a partition by how much of a node you need:

  • GPU: whole nodes, all eight GPUs on each (16 on the DGX-2), charged for the full node.
  • GPU-shared: 1 to 4 GPUs on a single node shared with other jobs, charged only for the GPUs you request.
  • GPU-small: interactive and small or quick jobs on a single V100-32 node.

For submission commands, sample scripts, and node-sharing details, see the Bridges-2 Batch Jobs guide. Partition-specific information is on the GPU and GPU-Shared partitions section of the user guide.

Queue specifications

Metrics updated 2026-09-29

Queue CPU cores / node GPUs / node Num nodes Node RAM Max wallclock Wait time
30-day trend
 Wall time
30-day trend
 Number of jobs run
30 days
GPU-shared Varies by node type Varies by node type Varies by node type Varies by node type Varies by node type
GPU-shared wait time: average 25.9 hours, range 0.2 to 96.1 hours over 30 days
GPU-shared wall time: average 2.2 hours, range 0.3 to 3.5 hours over 30 days, wall-time limit 48h
64,045
V100-32 Node. Jobs using 1 to 4 of a node's V100-32 GPUs, shared with other jobs; charged per GPU. 2x Intel Xeon Gold 6248 "Cascade Lake" (40 cores) 8 NVIDIA Tesla V100-32GB SXM2 (32 GB vRAM) 23 512 GB 48h See queue heading See queue heading See queue heading
H100-80 Node. Jobs using 1 to 4 of a node's H100-80 GPUs, shared with other jobs; charged per GPU. 2x Intel Xeon "Sapphire Rapids" 8470 (104 cores) 8 NVIDIA H100-80GB SXM5 (80 GB vRAM) 10 2 TB 48h See queue heading See queue heading See queue heading
V100-16 Node. Jobs using 1 to 4 of a node's V100-16 GPUs, shared with other jobs; charged per GPU. 2x Intel Xeon Gold 6148 (40 cores) 8 NVIDIA V100-16GB (16 GB vRAM) 9 192 GB 48h See queue heading See queue heading See queue heading
L40S-48 Node. Jobs using 1 to 4 of a node's L40S-48 GPUs, shared with other jobs; charged per GPU. 2x Intel Xeon 6740E (192 cores) 8 NVIDIA L40S-48GB (48 GB vRAM) 3 1 TB 48h See queue heading See queue heading See queue heading
GPU Varies by node type Varies by node type Varies by node type Varies by node type Varies by node type
GPU wait time: average 29.5 hours, range 0 to 270.9 hours over 30 days
GPU wall time: average 4.1 hours, range 0.5 to 14.1 hours over 30 days, wall-time limit 48h
553
V100-32 Node. Jobs using one or more whole V100-32 nodes, all eight GPUs each; charged for the full node. 2x Intel Xeon Gold 6248 "Cascade Lake" (40 cores) 8 NVIDIA Tesla V100-32GB SXM2 (32 GB vRAM) 23 512 GB 48h See queue heading See queue heading See queue heading
H100-80 Node. Jobs using one or more whole H100-80 nodes, all eight GPUs each; charged for the full node. 2x Intel Xeon "Sapphire Rapids" 8470 (104 cores) 8 NVIDIA H100-80GB SXM5 (80 GB vRAM) 10 2 TB 48h See queue heading See queue heading See queue heading
V100-16 Node. Jobs using one or more whole V100-16 nodes, all eight GPUs each; charged for the full node. 2x Intel Xeon Gold 6148 (40 cores) 8 NVIDIA V100-16GB (16 GB vRAM) 9 192 GB 48h See queue heading See queue heading See queue heading
L40S-48 Node. Jobs using one or more whole L40S-48 nodes, all eight GPUs each; charged for the full node. 2x Intel Xeon 6740E (192 cores) 8 NVIDIA L40S-48GB (48 GB vRAM) 3 1 TB 48h See queue heading See queue heading See queue heading
DGX-2 special node. Jobs using the whole DGX-2 node and all 16 V100-32 GPUs; charged for the full node. 2x Intel Xeon Platinum 8168 (48 cores) 16 NVIDIA Volta V100-32GB (32 GB vRAM) 1 1500 GB 48h See queue heading See queue heading See queue heading
GPU-small
Used for interactive and small or quick GPU jobs on a single V100-32 node.
 2x Intel Xeon Gold 6248 (40 cores) 8 NVIDIA Tesla V100-32GB SXM2 (32 GB vRAM) 1 512 GB 48h
GPU-small wait time: average 1.9 hours, range 0 to 3.1 hours over 30 days
GPU-small wall time: average 1.6 hours, range 0 to 2 hours over 30 days, wall-time limit 48h
96

Software Documentation

The following software packages are among the most frequently used on Bridges-2 GPU, based on job data from XDMoD.

Most Frequently Used

Application Description Research Discipline Jobs
python Python is a high-level, interpreted programming language known for its simplicity and readability. It supports multiple programming paradigms and has a vast ecosystem of libraries and frameworks. Computer & Information Sciences, Software Engineering, Systems & Development 175,529
amber Amber is a suite of highly extensible molecular simulation programs. It is designed for simulations of biomolecules such as proteins, nucleic acids, and carbohydrates, and can also be used for small molecules. Biological Sciences 4,243
q-espresso Quantum ESPRESSO is an integrated suite of computer codes for electronic-structure calculations and materials modeling at the nanoscale. Condensed Matter Physics 2,862
gromacs GROMACS (GROningen MAssive Parallel MD for Molecular Dynamics) is a versatile package for molecular dynamics simulations with a strong emphasis on high-performance computing capabilities. Biological Sciences 1,008
namd NAMD (NAnoscale Molecular Dynamics) is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. It is optimized for the simulation of biomolecular systems containing millions of atoms. Biochemistry and Molecular Biology 583
lammps LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator) is a classical molecular dynamics code designed for simulating large-scale atomistic systems. It is highly versatile and can be used to model a wide range of materials and complex molecular structures. Chemical Sciences 272
r R is a free software environment for statistical computing and graphics. It compiles and runs on a wide variety of UNIX platforms, Windows, and MacOS. Computer Science 181
athena 164
pegasus 93
castro 72

SEE ALL SOFTWARE AVAILABLE ON BRIDGES-2 GPU

Storage Documentation

Bridges-2 has two persistent spaces: a small Home for code and configuration, and the Ocean project space for active data and results. There is no separate scratch tier; node-local disk and RAM disk last only for the duration of a job. Ocean is shared across all of Bridges-2. See the File Spaces section of the user guide for more information.

File System

Directory Path Quota Purge Backup Notes
Home $HOME 25 GB Purged 3 months after allocation expires Backed up daily
Projects $PROJECT See notes Purged 3 months after allocation expires Not backed up Quota size depends on allocation
Node-local $LOCAL See notes Purged at job end Not backed up Quota varies by node type
Node memory $RAMDISK See notes Purged at job end Not backed up Quota varies by node type

File Transfer Documentation

Globus is recommended for large or many-file transfers, since it retries and resumes automatically. Use rsync, scp, or sftp for command-line transfers. Run all transfers through the Data Transfer Node (data.bridges2.psc.edu), not the login nodes.

For more information, please visit the Transferring Files section of the user guide.

Supported Methods Data Transfer Node / Globus Collection Notes
GLOBUS | RECOMMENDED PSC Bridges-2 /ocean and /jet filesystems https://app.globus.org
RSYNC data.bridges2.psc.edu
SCP data.bridges2.psc.edu
SFTP data.bridges2.psc.edu

Datasets Documentation

Name Description
2019nCoVR

COVID-19 genomic surveillance data and metadata (hosted by NGDC).

2019nCoVR Webpage
AlphaFold

Predicted protein structures for the human proteome and other key proteins.

AlphaFold Webpage
Path in Bridges-2: /ocean/datasets/community/alphafold
CIFAR-10

60,000 labeled images across 10 classes; standard image-classification benchmark.

CIFAR-10 Webpage

Path in Bridges-2: /ocean/datasets/community/cifar
COCO

Large-scale image dataset for object detection, segmentation, and captioning.

COCO Dataset Webpage
Path in Bridges-2: /ocean/datasets/community/COCO
CosmoFlow

~10,000 cosmological dark-matter simulations. Access requires a request via the CosmoFlow request form.

CosmoFlow Webpage
Path in Bridges-2: /ocean/datasets/community/cosmoflow
ImageNet

Image dataset organized by WordNet hierarchy.

ImageNet Webpage
Path in Bridges-2: /ocean/datasets/community/imagenet
MNIST

Classic handwritten-digit dataset for image-processing benchmarks.

Path in Bridges-2: /ocean/datasets/community/mnist
Natural Language Tool Kit Data

Corpora, grammars, and trained models for NLP.

NLTK Data Webpage
Path in Bridges-2: /ocean/datasets/community/nltk
OpenWebText

Path in Bridges-2: /ocean/datasets/community/openwebtext
PREVENT-AD

Longitudinal multimodal data from cognitively healthy older adults at risk for Alzheimer's, from two prevention trials.

Path in Bridges-2: /ocean/datasets/community/prevent_ad
TCGA Images

Path in Bridges-2: /ocean/datasets/community/tcga_images
Genomics datasets

These datasets are available to anyone with an allocation on Bridges-2. They are stored under /ocean/datasets/community/genomics.

AUGUSTUS, BLAST, CheckM, Dammit, Homer, Kraken2, Pfam, Prokka, Repbase