Unclassified Systems
Barfoot is an HPE Cray EX4000 system located at the ERDC DSRC. It has 1,092 standard compute nodes, 4 large-memory nodes, and 12 1-GPU visualization nodes (a total of 212,736 compute cores). It has 382 TB of usable memory and is rated at 8.2 peak PFLOPS.
| Login | Standard | Large-Memory | Visualization | |
|---|---|---|---|---|
| Total Nodes | 10 | 1,092 | 4 | 12 |
| Processor | AMD 9654 Genoa | AMD 9654 Genoa | AMD 9654 Genoa | AMD 9654 Genoa |
| Processor Speed | 2.4 GHz | 2.4 GHz | 2.4 GHz | 2.4 GHz |
| Sockets / Node | 2 | 2 | 2 | 2 |
| Cores / Node | 192 | 192 | 192 | 192 |
| Total CPU Cores | 1,920 | 209,664 | 768 | 2,304 |
| Usable Memory / Node | 8 GB | 335 GB | 2,050 GB | 700 GB |
| Accelerators / Node | None | None | None | 1 |
| Accelerator | N/A | N/A | N/A | NVIDIA L40S PCIe 5 |
| Memory / Accelerator | N/A | N/A | N/A | 48 GB |
| Storage on Node | 985 GB NVMe SSD | None | 10 TB NVMe SSD | None |
| Interconnect | Ethernet | HPE Slingshot | HPE Slingshot | HPE Slingshot |
| Operating System | SLES 15 | SLES 15 | SLES 15 | SLES 15 |
| Priority | Queue Name | Max Wall Clock Time | Max Cores Per Job | Description |
|---|---|---|---|---|
| Highest | urgent | 168 Hours | 9,408 | Jobs belonging to DoD HPCMP Urgent Projects |
|
debug^ | 1 Hour | 13,824 | Time/resource-limited for user testing and debug purposes |
| HIE | 24 Hours | 192 | Rapid response for interactive work. For more information see the HPC Interactive Environment (HIE) User Guide. | |
| high_lw | 168 Hours | 7,488 | Long-walltime jobs belonging to DoD HPCMP High Priority Projects | |
| high_lg | 24 Hours | 100,032 | Large jobs belonging to DoD HPCMP High Priority Projects | |
| frontier_lw | 168 Hours | 7,488 | Long-walltime jobs belonging to DoD HPCMP Frontier Projects | |
| frontier_lg | 24 Hours | 100,032 | Large jobs belonging to DoD HPCMP Frontier Projects | |
| standard_lw | 168 Hours | 7,488 | Long-walltime standard jobs | |
| standard_lg | 24 Hours | 100,032 | Large standard jobs | |
| serial | 168 Hours | 1 | Single-core serial jobs. 1 core per hour charged to project allocation. | |
| transfer | 48 Hours | 1 | Data transfer for user jobs. Not charged against project allocation. See the ERDC DSRC Archive Guide, section 5.2. | |
| Lowest | background^^ | 4 Hours | 9,408 | User jobs that are not charged against the project allocation |
^ The running job limit on the debug queue per user is 2.
^^ The running job limit on the background queue per user is 3.
Blueback is an HPE Cray EX4000 system located at the Navy DSRC. It has 1,280 standard compute nodes, 16 large-memory nodes, 32 AI/ML nodes, and 24 visualization nodes (a total of 1,352 compute nodes or 256,512 compute cores).
Available Documentation
| Date / Time | Details |
|---|---|
| 2026 Sep 15 10:00 - 18:00 CT | Network Maintenance |
| 2026 Sep 22 09:00 - 17:00 CT | Software |
| Login | Standard | Large-Memory | Visualization | AI/ML | Transfer | |
|---|---|---|---|---|---|---|
| Total Nodes | 12 | 1,280 | 16 | 24 | 32 | 6 |
| Processor | AMD 9654 Genoa | AMD 9654 Genoa | AMD 9654 Genoa | AMD 9654 Genoa | AMD Instinct MI300A APU | AMD 9654 Genoa |
| Processor Speed | 2.4 GHz | 2.4 GHz | 2.4 GHz | 2.4 GHz | 2.4 GHz | 2.4 GHz |
| Sockets / Node | 2 | 2 | 2 | 2 | 2 | 2 |
| Cores / Node | 192 | 192 | 192 | 192 | 96 | 192 |
| Total CPU Cores | 2,304 | 245,760 | 3,072 | 4,608 | 3,072 | 1,152 |
| Usable Memory / Node | 719 GB | 360 GB | 1.4 TB | 733 GB | 478 GB | 703 GB |
| Accelerators / Node | None | None | None | 1 | 4 | None |
| Accelerator | N/A | N/A | N/A | NVIDIA L40 | NVIDIA L40 | N/A |
| Memory / Accelerator | N/A | N/A | N/A | 48 GB | 128 GB | N/A |
| Storage on Node | None | None | 3.2 TB NVMe | None | 2 x 3.84 TB NVMe | None |
| Interconnect | HPE Slingshot | HPE Slingshot | HPE Slingshot | HPE Slingshot | HPE Slingshot | HPE Slingshot |
| Operating System | SLES | SLES | SLES | SLES | SLES | SLES |
| Priority | Queue Name | Max Wall Clock Time | Max Cores Per Job | Description |
|---|---|---|---|---|
| Highest | urgent | 24 Hours | 24,756 | Jobs belonging to DoD HPCMP Urgent Projects |
|
|
debug | 30 Minutes | 16,128 | Time/resource-limited for user testing and debug purposes |
| HIE | 24 Hours | 4,608 | Rapid response for interactive work. For more information see the HPC Interactive Environment (HIE) User Guide. | |
| frontier | 168 Hours | 98,304 | Jobs belonging to DoD HPCMP Frontier Projects | |
| high | 168 Hours | 98,304 | Jobs belonging to DoD HPCMP High Priority Projects | |
| serial | 168 Hours | 1 | Single-core serial jobs.1 core per hour charged to project allocation. | |
| standard | 168 Hours | 49,152 | Standard jobs | |
| transfer | 48 Hours | 1 | Data transfer for user jobs. Not charged against project allocation. See the Navy DSRC Archive Guide, section 5.2. | |
| Lowest | background | 4 Hours | 6,144 | User jobs that are not charged against the project allocation |
Carpenter is an HPE Cray EX4000 system located at the ERDC DSRC. It has 1,632 standard compute nodes, 4 large-memory nodes, and 8 GPU nodes (a total of 313,344 compute cores). It has 585 TB of useable memory and is rated at 17.65 peak PFLOPS.
| Login | Standard | Large-Memory | Visualization | |
|---|---|---|---|---|
| Total Nodes | 10 | 1,632 | 4 | 8 |
| Processor | AMD 9654 Genoa | AMD 9654 Genoa | AMD 9654 Genoa | AMD 7713 Milan |
| Processor Speed | 2.4 GHz | 2.4 GHz | 2.4 GHz | 2.0 GHz |
| Sockets / Node | 2 | 2 | 2 | 2 |
| Cores / Node | 192 | 192 | 192 | 128 |
| Total CPU Cores | 1,920 | 313,344 | 768 | 1,024 |
| Usable Memory / Node | 8 GB | 349 GB | 2.973 TB | 467 GB |
| Accelerators / Node | None | None | None | 1 |
| Accelerator | N/A | N/A | N/A | NVIDIA A40 PCIe 4 |
| Memory / Accelerator | N/A | N/A | N/A | 48 GB |
| Storage on Node | 1.3 TB NVMe SSD | None | 8.8 TB NVMe SSD | None |
| Interconnect | Ethernet | HPE Slingshot | HPE Slingshot | HPE Slingshot |
| Operating System | SLES 15 | SLES 15 | SLES 15 | SLES 15 |
| Priority | Queue Name | Max Wall Clock Time | Max Cores Per Job | Max Queued Per User | Max Running Per User | Description |
|---|---|---|---|---|---|---|
| Highest | urgent | 24 Hours | 9,408 | N/A | N/A | Jobs belonging to DoD HPCMP Urgent Projects |
|
|
debug^ | 1 Hour | 13,824 | N/A | 2 | Time/resource-limited for user testing and debug purposes |
| HIE | 24 Hours | 192 | 1 | 1 | Rapid response for interactive work. For more information see the HPC Interactive Environment (HIE) User Guide. | |
| high_lw | 168 Hours | 7,488 | N/A | 3 | Long-walltime jobs belonging to DoD HPCMP High Priority Projects | |
| high_lg | 24 Hours | 100,032 | N/A | 2 | Large jobs belonging to DoD HPCMP High Priority Projects | |
| frontier_lw | 168 Hours | 7,488 | N/A | 3 | Long-walltime jobs belonging to DoD HPCMP Frontier Projects | |
| frontier_lg | 24 Hours | 100,032 | N/A | 2 | Large jobs belonging to DoD HPCMP Frontier Projects | |
| standard_lw | 168 Hours | 7,488 | N/A | 3 | Long-walltime standard jobs | |
| standard_lg | 24 Hours | 100,032 | N/A | 2 | Large standard jobs | |
| serial | 168 Hours | 1 | N/A | 10 | Single-core serial jobs. 1 core per hour charged to project allocation. | |
| transfer | 48 Hours | 1 | N/A | 10 | Data transfer for user jobs. Not charged against project allocation. See the ERDC DSRC Archive Guide, section 5.2. | |
| Lowest | background^^ | 4 Hours | 9,408 | N/A | 3 | User jobs that are not charged against the project allocation |
^ The running job limit on the debug queue per user is 2.
^^ The running job limit on the background queue per user is 3.
Fran is an HPE Cray EX4000 system located at the ARL DSRC. It has 792 standard compute nodes, 24 large-memory nodes, 24 visualization nodes, 32 AI/ML nodes and 30 PSF (Persistent System Framework) nodes (a total of 173,760 compute cores). It has 434 TB of usable memory and is rated at 8.826 peak PFLOPS.
| Login | Standard | Large Memory | Visualization | AI/ML | Transfer | PSF^ | |
|---|---|---|---|---|---|---|---|
| Total Nodes | 12 | 792 | 24 | 24 | 32 | 3 | 30 |
| Processor | AMD 9654 Genoa | AMD 9654 Genoa | AMD 9654 Genoa | AMD 9654 Genoa | AMD 9654 Genoa | AMD 9654 Genoa | AMD 9654 Genoa |
| Processor Speed | 2.4 GHz | 2.4 GHz | 2.4 GHz | 2.4 GHz | 2.4 GHz | 2.4 GHz | 2.4 GHz |
| Sockets / Node | 2 | 2 | 2 | 2 | 2 | 2 | 2 |
| Cores / Node | 192 | 192 | 192 | 192 | 192 | 192 | 192 |
| Total CPU Cores | 2304 | 152064 | 4608 | 4608 | 6144 | 576 | 5760 |
| Usable Memory / Node | 385 GB | 385 GB | 2.36 TB | 768 GB | 768 GB | 768 GB | 768 GB |
| Accelerators / Node | None | None | None | 1 | 2 | None | None |
| Accelerator | n/a | n/a | n/a | NVIDIA L40S | NVIDIA H100 NVL | n/a | n/a |
| Memory / Accelerator | n/a | n/a | n/a | 48 GB | 141 GB | n/a | n/a |
| Storage on Node | 1.2 TB | None | None | None | None | None | |
| Interconnect | Ethernet | HPE Slingshot | HPE Slingshot | HPE Slingshot | HPE Slingshot | HPE Slingshot | HPE Slingshot |
| Operating System | RHEL9 | RHEL9 | RHEL9 | RHEL9 | RHEL9 | RHEL9 | RHEL9 |
^ Persistent Services Framework (PSF) provides functionality to bridge the gap of support that exists for DoD users who are interested in leveraging HPC resources but require access to persistent services as a component of their end-to-end workflow. Please reach out to ARL DSRC Helpdesk (dsrchelp@arl.hpc.mil) to request PSF resources.
| Priority | Queue Name | Max Wall Clock Time | Max Cores Per Job | Max Queued Per User | Max Running Per User | Description |
|---|---|---|---|---|---|---|
| Highest | urgent | 96 Hours | 162,816 | N/A | N/A | Jobs belonging to DoD HPCMP Urgent Projects |
|
|
HIE | 24 Hours | 4,608 | N/A | N/A | Rapid response for interactive work. For more information see the HPC Interactive Environment (HIE) User Guide. |
| transfer | 48 Hours | 576 | N/A | 3 | Data transfer for user jobs. Not charged against project allocation. See the ARL DSRC Archive Guide, section 5.2. | |
| debug^ | 1 Hour | 162,816 | N/A | N/A | Time/resource-limited for user testing and debug purposes | |
| AI/ML | 168 Hours | 6,144 | N/A | N/A | Jobs requesting AI/ML nodes. | |
| high | 168 Hours | 162,816 | N/A | 100,000 | Jobs belonging to DoD HPCMP High Priority Projects | |
| frontier | 168 Hours | 162,816 | N/A | 42 | Jobs belonging to DoD HPCMP Frontier Projects | |
| cots | 96 Hours | 162,816 | N/A | 20 | Jobs belonging to DoD HPCMP Frontier Projects | |
| interactive | 12 Hours | 162,816 | N/A | 3 | Rapid response for interactive work. | |
| standard | 168 Hours | 162,816 | N/A | N/A | Standard jobs | |
| standard-long | 336 Hours | 162,816 | N/A | 2 | Standard jobs needing additional wall time | |
| Lowest | background^^ | 24 Hours | 162,816 | N/A | 2 | User jobs that are not charged against the project allocation |
^ The running job limit on the debug queue per user is 4.
^^ The running job limit on the background queue per user is 1.
Jean is a Liqid system located at the ARL DSRC. It has 488 standard compute nodes, 20 large-memory nodes, 42 AI/ML nodes, and 26 visualization nodes.
| Login | Standard | Large-Memory | Visualization | AI/ML | Transfer | |
|---|---|---|---|---|---|---|
| Total Nodes | 9 | 488 | 20 | 26 | 42 | 2 |
| Processor | Intel 9242 Cascade Lake | Intel 9242 Cascade Lake | Intel 9242 Cascade Lake | Intel 9242 Cascade Lake | Intel 9242 Cascade Lake | Intel 9242 Cascade Lake |
| Processor Speed | 2.3 GHz | 2.3 GHz | 2.3 GHz | 2.3 GHz | 2.3 GHz | 2.3 GHz |
| Sockets / Node | 2 | 2 | 2 | 2 | 2 | 2 |
| Cores / Node | 96^ | 96^ | 96^ | 96^ | 96^ | 96^ |
| Total CPU Cores | 864 | 47,242 | 2,304 | 2,304 | 2,784 | 192 |
| Usable Memory / Node | 748 GB | 385 GB or 757 GB | 1546 GB | 772 GB | 772 GB | 772 GB |
| Accelerators / Node | 1 or 4 | None | None | 1 | 4 | None |
| Accelerator | NVIDIA A100 PCIe4 | N/A | N/A | NVIDIA A100 PCIe 4 | NVIDIA A100 PCIe 4 | N/A |
| Memory / Accelerator | 40 GB | N/A | N/A | 40 GB | 40 GB | N/A |
| Storage on Node | 7 TB NVMe SSD | None | 14 TB NVMe SSD | 7 TB NVMe SSD | 7 TB NVMe SSD | N/A |
| Interconnect | HDR InfiniBand | HDR InfiniBand | HDR InfiniBand | HDR InfiniBand | HDR Infiniband | HDR InfiniBand |
| Operating System | RHEL 8 | RHEL 8 | RHEL 8 | RHEL 8 | RHEL 8 | RHEL 8 |
^ There are 96 cores/node. However, WEKA consumes 4 cores/node so only 92 cores/node are available. To ensure your job does not run slowly or encounter issues, 92 available cores/nodes should be used in your batch scripts. For example, if you want to run a 4-node job with all available cores (92 cores/node not 96 cores/node), the following Slurm directives should be utilized: --nodes=4, --ntasks=368 (not 384 cores) and --ntasks-per-node=92 (not 96 cores).
| Priority | Queue Name | Max Wall Clock Time | Max Cores Per Job | Max Queued Per User | Max Running Per User | Description |
|---|---|---|---|---|---|---|
| Highest | urgent | 96 Hours | 48,000^ | N/A | N/A | Jobs belonging to DoD HPCMP Urgent Projects |
|
|
HIE | 24 Hours | 2,496^ | N/A | N/A | Rapid response for interactive work. For more information see the HPC Interactive Environment (HIE) User Guide. |
| transfer | 48 Hours | 192^ | N/A | N/A | Data transfer for user jobs. Not charged against project allocation. See the ARL DSRC Archive Guide, section 5.2. | |
| debug^^ | 1 Hour | 54,408^ | N/A | N/A | Time/resource-limited for user testing and debug purposes | |
| high | 168 Hours | 48,000^ | N/A | N/A | Jobs belonging to DoD HPCMP High Priority Projects | |
| frontier | 168 Hours | 48,000^ | N/A | N/A | Jobs belonging to DoD HPCMP Frontier Projects | |
| cots | 96 Hours | 53,760^ | N/A | N/A | Jobs using COTS applications | |
| interactive | 12 Hours | 48,000^ | N/A | N/A | Rapid response for interactive work. | |
| standard | 168 Hours | 48,000^ | N/A | N/A | Standard jobs | |
| standard-long | 336 Hours | 48,000^ | N/A | N/A | Standard jobs needing additional wall time | |
| background^^^ | 4 Hours | 5,280^ | N/A | N/A | User jobs that are not charged against the project allocation | |
| AIML | 168 Hours | 4,032^ | N/A | N/A | Dedicated queue for jobs utilizing GPUs | |
| Lowest | serial | 96 Hours | 192^ | N/A | N/A | User jobs that are not charged against the project allocation |
^ There are 96 cores/node as reflected in the Max Cores per Job column in the table above. However, WEKA consumes 4 cores/node so only 92 cores/node are available. To ensure your job does not run slowly or encounter issues, 92 available cores/nodes should be used in your batch scripts. For example, if you want to run a 4-node job with all available cores (92 cores/node not 96 cores/node), the following Slurm directives should be utilized: --nodes=4, --ntasks=368 (not 384 cores) and --ntasks-per-node=92 (not 96 cores).
^^ The running job limit on the debug queue per user is 4.
^^^ The running job limit on the background queue per user is 1.
Makau is an HPE Cray XD2000 system located at the MHPCC DSRC. It has 10 mixed compute nodes with 1 Nvidia H100 GPU each and 16 AI/ML nodes with 4 Nvidia H100 GPUS each. It has 19.2TB of usable memory and is rated at 2.64 peak PFLOPS.
| Login | AI/ML | Mixed | Transfer | |
|---|---|---|---|---|
| Total Nodes | 4 | 16 | 10 | 2 |
| Processor | AMD EPYC 9534 | AMD EPYC 9534 | AMD EPYC 9534 | AMD EPYC 9534 |
| Processor Speed | 2.45 GHz | 2.45 GHz | 2.45 GHz | 2.45 GHz |
| Sockets / Node | 2 | 2 | 2 | 2 |
| Cores / Node | 128 | 128 | 128 | 128 |
| Total CPU Cores | 512 | 2,048 | 1,280 | 256 |
| Usable Memory / Node | 600 GB | 600 GB | 600 GB | 600 GB |
| Accelerators / Node | None | 4 | 1 | None |
| Accelerator | N/A | NVIDIA H100 SXM5 | NVIDIA H100 NVL | N/A |
| Memory / Accelerator | N/A | 80 GB HBM3 | 94 GB HBM3 | N/A |
| Storage on Node | None | 3.84 TB NVMe | 3.84 TB NVMe | None |
| Interconnect | HDR InfiniBand | HDR InfiniBand | HDR InfiniBand | HDR InfiniBand |
| Operating System | RHEL9 | RHEL9 | RHEL9 | RHEL9 |
| Priority | Queue Name | Max Wall Clock Time | Max Cores Per Job | Description |
|---|---|---|---|---|
| Highest | urgent | 24 Hours | 256 | Jobs belonging to DoD HPCMP Urgent Projects |
|
|
debug^ | 0.5 Hours | 128 | Time/resource-limited for user testing and debug purposes |
| high | 336 Hours | 1,536 | Jobs belonging to DoD HPCMP High Priority Projects | |
| frontier | 168 Hours | 1,024 | Jobs belonging to DoD HPCMP Frontier Projects | |
| standard | 168 Hours | 512 | Standard jobs | |
| background^^ | 4 Hours | 256 | User jobs that are not charged against the project allocation | |
| Lowest | transfer | 48 Hours | 256 | Data transfer for user jobs. Not charged against project allocation. See the AFRL DSRC Archive Guide, section 5.2. |
^ The running job limit on the debug queue per user is 1.
^^ The running job limit on the background queue per user is 1.
Narwhal is an HPE Cray EX system located at the Navy DSRC. It has 2,304 standard compute nodes, 26 large-memory nodes, 16 visualization accelerated nodes, 32 Single-GPU MLA accelerated nodes, and 32 Dual-GPU MLA accelerated nodes (a total of 2,410 compute nodes or 308,480 compute cores). It has 640 TB of memory and is rated at 13.5 peak PFLOPS.
Available Documentation
| Date / Time | Details |
|---|---|
| 2026 Sep 15 10:00 - 18:00 CT | Network Maintenance |
| Login | Standard | Large-Memory | Visualization | Single-GPU MLA | Dual-GPU MLA | |
|---|---|---|---|---|---|---|
| Total Nodes | 11 | 2,304 | 26 | 16 | 32 | 32 |
| Processor | AMD 7H12 Rome | AMD 7H12 Rome | AMD 7H12 Rome | AMD 7H12 Rome | AMD 7H12 Rome | AMD 7H12 Rome |
| Processor Speed | 2.6 GHz | 2.6 GHz | 2.6 GHz | 2.6 GHz | 2.6 GHz | 2.6 GHz |
| Sockets / Node | 2 | 2 | 2 | 2 | 2 | 2 |
| Cores / Node | 128 | 128 | 128 | 128 | 128 | 128 |
| Total CPU Cores | 1,408 | 294,912 | 3,328 | 2,048 | 4,096 | 4,096 |
| Usable Memory / Node | 226 GB | 238 GB | 995 GB | 234 GB | 239 GB | 239 GB |
| Accelerators / Node | None | None | None | 1 | 1 | 2 |
| Accelerator | N/A | N/A | N/A | NVIDIA V100 PCIe 3 | NVIDIA V100 PCIe 3 | NVIDIA V100 PCIe 3 |
| Memory / Accelerator | N/A | N/A | N/A | 32 GB | 32 GB | 32 GB |
| Storage on Node | 880 GB SSD | None | 1.8 TB SSD | None | 880 GB SSD | 880 GB SSD |
| Interconnect | HPE Slingshot | HPE Slingshot | HPE Slingshot | HPE Slingshot | HPE Slingshot | HPE Slingshot |
| Operating System | SLES | SLES | SLES | SLES | SLES | SLES |
| Priority | Queue Name | Max Wall Clock Time | Max Cores Per Job | Description |
|---|---|---|---|---|
| Highest | urgent | 24 Hours | 16,384 | Jobs belonging to DoD HPCMP Urgent Projects |
|
|
debug | 30 Minutes | 8,192 | Time/resource-limited for user testing and debug purposes |
| HIE | 24 Hours | 3,072 | Rapid response for interactive work. For more information see the HPC Interactive Environment (HIE) User Guide. | |
| frontier | 168 Hours | 16,384 | Jobs belonging to DoD HPCMP Frontier Projects | |
| high | 168 Hours | 16,384 | Jobs belonging to DoD HPCMP High Priority Projects | |
| serial | 168 Hours | 1 | Single-core serial jobs. 1 core per hour charged to project allocation. | |
| standard | 168 Hours | 16,384 | Standard jobs | |
| transfer | 48 Hours | 1 | Data transfer for user jobs. Not charged against project allocation. See the Navy DSRC Archive Guide, section 5.2. | |
| Lowest | background | 4 Hours | 8,192 | User jobs that are not charged against the project allocation |
Nautilus is a Penguin Computing TrueHPC system located at the Navy DSRC. It has 1,384 standard compute nodes, 16 large-memory nodes, 16 visualization accelerated nodes, 32 AI/ML nodes, and 32 High Core Performance nodes (a total of 1,480 compute nodes or 186,368 compute cores). It has 386 TB of memory and is rated at 8.5 peak PFLOPS.
Available Documentation
| Date / Time | Details |
|---|---|
| 2026 Sep 15 08:00 - 18:00 CT | System Maintenance |
| 2026 Sep 15 10:00 - 18:00 CT | Network Maintenance |
| 2026 Sep 23 06:00 - 17:00 CT | Facility |
| Login | Standard | Large-Memory | Visualization | AI/ML | High Core Performance | |
|---|---|---|---|---|---|---|
| Total Nodes | 14 | 1,384 | 16 | 16 | 32 | 32 |
| Processor | AMD 7713 Milan | AMD 7713 Milan | AMD 7713 Milan | AMD 7713 Milan | AMD 7713 Milan | AMD 73F3 Milan |
| Processor Speed | 2 GHz | 2 GHz | 2 GHz | 2 GHz | 2 GHz | 3.4 GHz |
| Sockets / Node | 2 | 2 | 2 | 2 | 2 | 2 |
| Cores / Node | 128 | 128 | 128 | 128 | 128 | 32 |
| Total CPU Cores | 1,792 | 166,912 | 2,048 | 2,048 / 16 | 4,096 / 128 | 1,024 |
| Usable Memory / Node | 433 GB | 231 GB | 998 GB | 491 GB | 491 GB | 491 GB |
| Accelerators / Node | None | None | None | 1 | 4 | None |
| Accelerator | N/A | N/A | N/A | NVIDIA A40 PCIe 4 | NVIDIA A100 SXM 4 | N/A |
| Memory / Accelerator | N/A | N/A | N/A | 48 GB | 40 GB | N/A |
| Storage on Node | 49 GB NVMe SSD | None | 49 GB NVMe SSD | None | 49 GB NVMe SSD | None |
| Interconnect | HDR InfiniBand | HDR InfiniBand | HDR InfiniBand | HDR InfiniBand | HDR InfiniBand | HDR InfiniBand |
| Operating System | RHEL | RHEL | RHEL | RHEL | RHEL | RHEL |
| Priority | Queue Name | Max Wall Clock Time | Max Cores Per Job | Max Queued Per User | Max Running Per User | Description |
|---|---|---|---|---|---|---|
| Highest | urgent | 24 Hours | 16,384 | N/A | 10 | Jobs belonging to DoD HPCMP Urgent Projects |
|
|
debug | 30 Minutes | 10,752 | N/A | 4 | Time/resource-limited for user testing and debug purposes |
| HIE | 24 Hours | 3,072 | N/A | 3 | Rapid response for interactive work. For more information see the HPC Interactive Environment (HIE) User Guide. | |
| frontier | 168 Hours | 65,536 | N/A | 50 | Jobs belonging to DoD HPCMP Frontier Projects | |
| high | 168 Hours | 65,536 | N/A | 50 | Jobs belonging to DoD HPCMP High Priority Projects | |
| serial | 168 Hours | 1 | N/A | NA | Single-core serial jobs. 1 core per hour charged to project allocation. | |
| standard | 168 Hours | 16,384 | N/A | 150 | Standard jobs | |
| transfer | 48 Hours | 1 | N/A | NA | Data transfer for user jobs. Not charged against project allocation. See the Navy DSRC Archive Guide, section 5.2. | |
| Lowest | background | 4 Hours | 4,096 | N/A | 10 | User jobs that are not charged against the project allocation |
Raider is a Penguin Computing TrueHPC system located at the AFRL DSRC. It has 1,480 standard compute nodes, 8 large-memory nodes, and 24 Visualization nodes, 32 MLA nodes, and 64 High Clock nodes (a total of 199,680 compute cores). It has 447 TB of memory and is rated at 9 peak PFLOPS.
Available Documentation
| Date / Time | Details |
|---|---|
| 2026 Sep 09 07:00 - Sep 10 07:00 ET (Completed) | Software |
| Login | Login-viz | Standard | Large-Memory | Visualization | MLA | High Clock | Transfer | |
|---|---|---|---|---|---|---|---|---|
| Total Nodes | 6 | 4 | 1,480 | 8 | 24 | 32 | 64 | 2 |
| Processor | AMD 7713 Milan | AMD 7713 Milan | AMD 7713 Milan | AMD 7713 Milan | AMD 7713 Milan | AMD 7713 Milan | AMD 73F3 Milan | AMD 7713 Milan |
| Processor Speed | 2.0 GHz | 2.0 GHz | 2.0 GHz | 2.0 GHz | 2.0 GHz | 2.0 GHz | 3.4 GHz | 2.0 GHz |
| Sockets / Node | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 |
| Cores / Node | 128 | 128 | 128 | 128 | 128 | 128 | 32 | 128 |
| Total CPU Cores | 768 | 512 | 189,440 | 1,024 | 3,072 | 4,096 | 2,048 | 256 |
| Usable Memory / Node | 503 GB | 503 GB | 251 GB | 2.0 TB | 503 GB | 503 GB | 503 GB | 503 GB |
| Accelerators / Node | 1 | 1 | None | None | 1 | 4 | None | None |
| Accelerator | NVIDIA A40 PCIe 4 | NVIDIA A100 SXM 4 | N/A | N/A | NVIDIA A40 PCIe 4 | NVIDIA A100 SXM 4 | N/A | N/A |
| Memory / Accelerator | 45 GB | 40 GB | N/A | N/A | 45 GB | 40 GB | N/A | N/A |
| Storage on Node | 960 GB NVMe SSD | 960 GB NVMe SSD | 1.91 TB NVMe SSD | 7.68 TB NVMe SSD | None | 3.84 TB NVMe SSD | None | None |
| Interconnect | HDR InfiniBand | HDR InfiniBand | HDR InfiniBand | HDR InfiniBand | HDR InfiniBand | HDR InfiniBand | HDR InfiniBand | HDR InfiniBand |
| Operating System | RHEL | RHEL | RHEL | RHEL | RHEL | RHEL | RHEL | RHEL |
| Priority | Queue Name | Max Wall Clock Time | Max Cores Per Job | Max Queued Per User | Max Running Per User | Description |
|---|---|---|---|---|---|---|
| Highest | urgent | 168 Hours | 92,160 | N/A | N/A | Jobs belonging to DoD HPCMP Urgent Projects |
|
|
debug | 1 Hour | 3,840 | 15 | 4 | Time/resource-limited for user testing and debug purposes |
| high | 168 Hours | 92,160 | N/A | N/A | Jobs belonging to DoD HPCMP High Priority Projects | |
| frontier | 168 Hours | 92,160 | N/A | N/A | Jobs belonging to DoD HPCMP Frontier Projects | |
| standard | 168 Hours | 92,160 | N/A | N/A | Standard jobs | |
| HIE | 24 Hours | 256 | 2 | 2 | Rapid response for interactive work. For more information see the HPC Interactive Environment (HIE) User Guide. | |
| transfer | 48 Hours | 1 | N/A | 12 | Data transfer for user jobs. Not charged against project allocation. See the AFRL DSRC Archive Guide, section 5.2. | |
| Lowest | background | 24 Hours | 3,840 | 35 | 10 | User jobs that are not charged against the project allocation |
Ruth is an HPE Cray EX system located at the ARL DSRC. It has 640 standard compute nodes, 24 large-memory nodes, 16 AI/ML nodes, and 24 visualization nodes.
Available Documentation
| Date / Time | Details |
|---|---|
| 2026 Sep 09 07:00 - Sep 14 10:00 ET (In Progress) | Facility |
| Login | Standard | Large-Memory | High Performance | AI/ML | Visualization | Transfer | |
|---|---|---|---|---|---|---|---|
| Total Nodes | 12 | 640 | 24 | 64 | 16 | 24 | 4 |
| Processor | AMD 9654 Genoa | AMD 9654 Genoa | AMD 9654 Genoa | AMD 73F3 Milan | AMD 7A53 Trento | AMD 7713 Milan | AMD 9654 Genoa |
| Processor Speed | 2.4 GHz | 2.4 GHz | 2.4 GHz | 3.5 GHz | 2.0 GHz | 2.0 GHz | 2.4 GHz |
| Sockets / Node | 2 | 2 | 2 | 2 | 1 | 2 | 2 |
| Cores / Node | 192 | 192 | 192 | 32 | 64 | 128 | 64 |
| Total CPU Cores | 2,304 | 122,880 | 4,608 | 2,048 | 1,024 | 3,072 | 256 |
| Usable Memory / Node | 780 GB | 360 GB | 1,500 GB | 500 GB | 500 GB | 500 GB | 780 GB |
| Accelerators / Node | None | None | None | None | 8 | 1 | None |
| Accelerator | N/A | N/A | N/A | N/A | AMD MI250X Infinity | NVIDIA A40 PCIe 4 | N/A |
| Memory / Accelerator | N/A | N/A | N/A | N/A | 128 GB HBM2e Memory | 48 GB | N/A |
| Storage on Node | None | None | 18 TB NVMe SSD | None | 3.5 TB NVMe SSD | None | 35 TB NVMe SSD |
| Interconnect | Ethernet | HPE Slingshot | HPE Slingshot | HPE Slingshot | HPE Slingshot | HPE Slingshot | HPE Slingshot |
| Operating System | SLES 15 | SLES 15 | SLES 15 | SLES 15 | SLES 15 | SLES 15 | SLES 15 |
| Priority | Queue Name | Max Wall Clock Time | Max Cores Per Job | Max Queued Per User | Max Running Per User | Description |
|---|---|---|---|---|---|---|
| Highest | urgent | 24 Hours | N/A | N/A | N/A | Jobs belonging to DoD HPCMP Urgent Projects |
|
|
transfer | 48 Hours | N/A | N/A | 3 | Data transfer for user jobs. Not charged against project allocation. See the ARL DSRC Archive Guide, section 5.2. |
| debug | 1 Hour | 16,384 | N/A | 2 | Time/resource-limited for user testing and debug purposes | |
| high | 168 Hours | N/A | N/A | 100,000 | Jobs belonging to DoD HPCMP High Priority Projects | |
| frontier | 168 Hours | N/A | N/A | 42 | Jobs belonging to DoD HPCMP Frontier Projects | |
| cots | 96 Hours | N/A | N/A | 20 | Jobs belonging to DoD HPCMP Frontier Projects | |
| HIE | 24 Hours | N/A | N/A | 2 | Rapid response for interactive work. For more information see the HPC Interactive Environment (HIE) User Guide. | |
| interactive | 12 Hours | N/A | N/A | 3 | Rapid response for interactive work. | |
| standard | 168 Hours | 100,000 | N/A | N/A | Standard jobs | |
| standard-long | 200 Hours | 4,096 | N/A | 2 | Standard jobs needing additional wall time | |
| Lowest | background^ | 24 Hours | 10,000 | N/A | 2 | User jobs that are not charged against the project allocation |
^ The running job limit on the background queue is 10.
Wheat is a Liqid system located at the ERDC DSRC. It has 770 standard compute nodes, 7 large-memory nodes, 16 1-GPU visualization nodes, 63 4-GPU MLA nodes, and 33 6-GPU MLA nodes (a total of 81,502 compute cores). It has 376 TB of useable memory and is rated at 4.2 PFLOPS.
Available Documentation
| Date / Time | Details |
|---|---|
| 2026 Oct 01 08:00 - Oct 09 17:00 CT | Software |
| Login | Standard | Large-Memory | Visualization | 4-GPU MLA | 6-GPU MLA | |
|---|---|---|---|---|---|---|
| Total Nodes | 6 | 770 | 7 | 16 | 63 | 33 |
| Processor | Intel 9242 Cascade Lake | Intel 9242 Cascade Lake | Intel 9242 Cascade Lake | Intel 9242 Cascade Lake | Intel 9242 Cascade Lake | Intel 9242 Cascade Lake |
| Processor Speed | 2.3 GHz | 2.3 GHz | 2.3 GHz | 2.3 GHz | 2.3 GHz | 2.3 GHz |
| Sockets / Node | 2 | 2 | 2 | 2 | 2 | 2 |
| Cores / Node | 92 | 92 | 92 | 92 | 92 | 92 |
| Total CPU Cores | 552 | 70,840 | 644 | 1,472 | 5,796 | 3,036 |
| Usable Memory / Node | 8 GB | 356 GB | 3 TB | 731 GB | 731 GB | 731 GB |
| Accelerators / Node | 1 | None | 4 | 1 | 4 | 6 |
| Accelerator | NVIDIA A100 | N/A | NVIDIA A100 | NVIDIA A100 | NVIDIA A100 | NVIDIA A100 |
| Memory / Accelerator | 40 GB | N/A | 40 GB | 40 GB | 40 GB | 40 GB |
| Storage on Node | None | None | None | None | None | None |
| Interconnect | Ethernet | InfiniBand | InfiniBand | InfiniBand | InfiniBand | InfiniBand |
| Operating System | RHEL 8 | RHEL 8 | RHEL 8 | RHEL 8 | RHEL 8 | RHEL 8 |
| Priority | Queue Name | Max Wall Clock Time | Max Cores Per Job | Description |
|---|---|---|---|---|
| Highest | urgent | 168 Hours | 2,392 | Jobs belonging to DoD HPCMP Urgent Projects |
|
|
debug^ | 1 Hour | 3,680 | Time/resource-limited for user testing and debug purposes |
| HIE | 24 Hours | 92 | Rapid response for interactive work. For more information see the HPC Interactive Environment (HIE) User Guide. | |
| high_lw | 168 Hours | 1,932 | Long-walltime jobs belonging to DoD HPCMP High Priority Projects | |
| high_lg | 24 Hours | 35,420 | Large jobs belonging to DoD HPCMP High Priority Projects | |
| frontier_lw | 168 Hours | 1,932 | Long-walltime jobs belonging to DoD HPCMP Frontier Projects | |
| frontier_lg | 24 Hours | 35,420 | Large jobs belonging to DoD HPCMP Frontier Projects | |
| standard_lw | 168 Hours | 1,932 | Long-walltime standard jobs | |
| standard_lg | 24 Hours | 35,420 | Large standard jobs | |
| transfer | 48 Hours | 1 | Data transfer for user jobs. Not charged against project allocation. See the ERDC DSRC Archive Guide, section 5.2. | |
| Lowest | background^^ | 4 Hours | 48 | User jobs that are not charged against the project allocation |
^ The running job limit on the debug queue per user is 2.
^^ The running job limit on the background queue per user is 3.