Anvil AI provides access to NVIDIA H100 GPUs for advanced AI workloads. It is designed for deep learning, large model training, and other compute-intensive AI applications.
Login to Anvil AI
- Users can use their ACCESS account to receive an allocation and login.
- Logging into your ACCESS account will require Duo two-factor authentication.
SSH Login
$ ssh x-ACCESS-username@anvil.rcac.purdue.edu
File Transfer
| Supported Methods | Data Transfer Node | URL |
|---|---|---|
| SCP | anvil.rcac.purdue.edu | |
| RSYNC | anvil.rcac.purdue.edu |
Storage
File System
| Directory | Path | Quota | Purge | Backup | Notes |
|---|---|---|---|---|---|
| Anvil ZFS | /home | 25 GB | Not Purged | Home directories: area for storing personal software, scripts, compiling, editing, etc. | |
| Anvil ZFS | /apps | N/A | Not purged | Applications | |
| Anvil GPFS | /anvil | N/A | Not purged | ||
| Anvil GPFS | /anvil/scratch | 100 TB | Files older than 30-day will be purged | User scratch: area for job I/O activity, temporary storage | |
| Anvil GPFS | /anvil/projects | 5 TB | Removed 90 days after allocation expiration | Per allocation: area for shared data in a project, common datasets and software installation | |
| Anvil GPFS | /anvil/datasets | Common data sets (not allocated to users) |
External Storage
Files in scratch directories are not recoverable. Files in scratch directories are not backed up. If you accidentally delete a file, a disk crashes, or old files are purged, they cannot be restored.
$PROJECTspace. The project space will be created for each allocation. $PROJECT and $WORK variables refer to the same location and can be used interchangeably.
ANVIL CEPH
Anvil Ceph is intended to provide scalable, fault-tolerant, and high-throughput storage for large or persistent research data. It supports both object and block storage, making it suitable for hosting shared datasets, storing long-term research outputs, and enabling data access for containerized or cloud-integrated workflows. Ceph complements the Lustre-based storage tiers by offering durable and easily expandable storage for diverse data management needs.
Inspecting file system quotas
To check the quota of different file systems, type myquota at the command line.
Jobs
Queue specifications
| Name | Purpose | CPUs | GPUs | RAM | Jobs
30 days
|
Wait Time
30-day trend
|
Wall Time
30-day trend
|
|---|---|---|---|---|---|---|---|
| ai | 96 Intel Xeon Platinum 8468 CPUs | 4 Nvidia H100 GPUs | 1TB | — | — | — |
Datasets
| Name | Description |
|---|---|
| AI | https://datasetdocs.readthedocs.io/en/latest/ai/index.html |
| Covariates | https://datasetdocs.readthedocs.io/en/latest/Covariates/index.html |
| Geospatial | https://datasetdocs.readthedocs.io/en/latest/geospatial/index.html |
| Hydrological | https://datasetdocs.readthedocs.io/en/latest/hydrological/index.html |
| Genomes | https://datasetdocs.readthedocs.io/en/latest/igenomes/index.html |
| Meteorological | https://datasetdocs.readthedocs.io/en/latest/meteorological/index.html |
| GeoAI | https://datasetdocs.readthedocs.io/en/latest/geoai/index.html |