Delta CPU

2FA/MFA RP account needed

Delta CPU is the CPU-only portion of NCSA’s Delta system, an ACCESS-allocated high-performance computing resource designed for workloads that do not require GPU acceleration. Delta CPU is well suited for CPU-based computation, parallel computation, data analysis, software compilation, and pre/post-processing tasks.

Delta CPU nodes are dual-socket AMD EPYC 7763 “Milan” compute nodes with 128 CPU cores, 256 GB RAM, and 0.74 TB of local storage per node. CPU jobs run through the Slurm scheduler and can use Delta’s shared file systems, command-line batch environment, and Open OnDemand interfaces.

GET AN ACCOUNT ON DELTA CPU

NCSA Project Management

Jobs

Jobs on Delta CPU resources are submitted through the Slurm scheduler. Users can submit batch jobs with sbatch, start interactive jobs with srun or salloc, or access compute resources through Open OnDemand applications such as JupyterLab, VS Code Code Server, and noVNC Desktop.

Delta CPU jobs should use the CPU partitions: cpu for standard production jobs, cpu-interactive for short interactive sessions, and cpu-preempt for checkpointing-capable jobs that can tolerate interruption. The standard cpu and cpu-preempt partitions allow jobs up to 48 hours, while the cpu-interactive partition has a 1-hour maximum duration.

Node sharing is the default behavior for Delta jobs. Users who need exclusive access to a CPU node can request all consumable resources for the node or use Slurm options such as --exclusive --mem=0. CPU and GPU allocations use separate local charge accounts, so CPU jobs should be submitted using the appropriate Delta CPU account.

For more information about running jobs on Delta, see the NCSA Delta [Running Jobs] documentation.

Queue specifications

Metrics updated 2026-06-16

Name Purpose CPUs GPUs RAM Jobs
30 days
Wait Time
30-day trend
Wall Time
30-day trend
cpu Standard CPU queue for production batch jobs 128 256GB 451,657
cpu wait time: average 28.3 hours, range 0.1 to 93.1 hours over 30 days 28.3
cpu wall time: average 0.5 hours, range 0 to 1.6 hours over 30 days 0.5
cpu-interactive Interactive CPU sessions for development, testing, debugging, compiling, and short interactive work 128 256GB 1,939
cpu-interactive wait time: average 0.5 hours, range 0 to 6.7 hours over 30 days 0.5
cpu-interactive wall time: average 0.3 hours, range 0.1 to 0.6 hours over 30 days 0.3
cpu-preempt Preemptible CPU jobs for checkpointing-capable, flexible, lower-priority workloads 128 256GB 88
cpu-preempt wait time: average 0.0 hours, range 0 to 0.3 hours over 30 days 0.0
cpu-preempt wall time: average 1.0 hours, range 0 to 4.5 hours over 30 days 1.0

Software

No software usage data is currently reported for Delta CPU in XDMoD.


Datasets

Name Description
ColabFold

ColabFold predicts protein folding structures and complexes using MMseqs2, which gives faster performance than AlphaFold. The ColabFold databases are available on Delta at /work/hdd/data/colabfold. The colabfold module may be loaded to set the COLABFOLD_PATH environment variable to this path.

ImageNet

ImageNet is an image database organized according to the WordNet hierarchy, where each node of the hierarchy is depicted by hundreds to thousands of images. ImageNet database(s), version ILSVRC 2012–2017, are available on Delta at /work/hdd/data/imagenet. The imagenet module may be loaded to set the IMAGENET_PATH environment variable to this path. Access requires membership in the delta_imagenet group.

alphafold

Protein structure prediction datasets used for bioinformatics research with AlphaFold.

models(ollama)

Pretrained large language models used with Ollama for local inference and AI experimentation.


Storage

File System

Directory Path Quota Purge Backup Notes
HOME /u 100 GB; 750,000 files Not purged 30 days Area for software, scripts, job files, and so on. Not intended as a source/destination for I/O during jobs.
PROJECTS /projects 500 GB. Up to 1-25 TB by allocation request. Not purged No Area for shared data for a project, common data sets, software, results, and so on.
WORK-HDD /work/hdd 1000 GB. Up to 1-100 TB by allocation request. Not purged No Area for computation, largest allocations, where I/O from jobs should occur. (this is now your scratch volume)
WORK-NVME /work/nvme NVME space is available upon request. Not purged No Area for computation, NVME is best for lots of small I/O from jobs.
TMP /tmp 0.74 TB (CPU) or 1.5 TB (GPU). Shared or dedicated depending on node usage by jobs. Purged after each job No Locally attached disk for fast small file I/O

File Transfer

Delta supports several methods for moving data to and from the system. Users may use scp or rsync for small to modest transfers, while Globus is recommended for large data transfers. GUI applications that use SSH, SCP, or SFTP must support Duo multi-factor authentication.

Files can also be shared externally from the /projects file system using Globus. To share data, users should create a sharing directory under their project space, move or copy the data into that directory, and connect to the “ACCESS Delta” Globus endpoint to create the shared collection.

The Delta Globus endpoint provides access to Delta home directories and shared project/work file systems, including /work/hdd and /work/nvme

For more information, see the NCSA Delta [File Sharing] and [Transferring Data] documentation.

Supported Methods Data Transfer Node URL
GLOBUS | RECOMMENDED https://app.globus.org/
SCP dt-login[01-04].delta.ncsa.illinois.edu
RSYNC dt-login[01-04].delta.ncsa.illinois.edu