Open Research Service
Welcome to the DoD HPCMP Open Research Service (ORS). The Open Research Service provides for computationally based science and engineering in support of the DoD. Systems are provided for high performance computing and utility computing. These systems operate entirely at the Unclassified level, with neither sensitive nor classified data or applications.
Obtaining an ORS Account
The process of getting an ORS account begins with getting an account on the HPCMP Portal to the Information Environment, commonly called a "pIE Account." If you do not yet have a pIE User Account, please visit Obtaining An Account and follow the instructions there. If you need assistance with any part of this process, please contact the HPC Help Desk at firstname.lastname@example.org. If you have an active pIE account, you may proceed to the next steps.
HPCMP Open Research Service Account Request Form
Directions: This form is required for a user to access an HPCMP Open Research Service (ORS) resource, whether that user is an existing or new HPCMP user. As with any other Open Research system that the program operates, all computational work performed on these resources must be either in the public domain or planned for the public domain by the nature of the computational work to be performed, generally basic research. The AWS GovCloud resource is protected up to Impact level 2 under the Information and Protection Level data protection protocol. Eligible users for the HPCMP Open Research Service resources must either be current HPCMP users who have used the HPCMP Open Research System, Copper, during FY 2019/20, or new users sponsored by a DoD RDT&E organization to perform computational work on open systems. Please note that HPCMP users who are eligible to use other unclassified HPCMP computational resources are NOT eligible to use the HPCMP Open Research Service resources.
New and existing HPCMP users must complete this form, providing user-specific information. Users should provide details of the computational work planned for the HPCMP resource, including their individual allocation request in core-hours. When completed, click submit.
Please Note: All requests will take a minimum of 24 hours to process once received. Request fulfillment time will vary based on the specific request and lead time required. For questions, please e-mail email@example.com.
* Denotes Required Fields
User Vetting Process
If you are new to the HPCMP and only need access to the Open Research Service, once you have a pIE account, you must complete the following:
US Citizens must provide copies of one of the following:
- Naturalization Certificate
- Birth Certificate
- US Passport Card
Foreign Nationals must provide copies of the following:
- Green Card Holders - Passport and Permanent Residence Card
- Exchange Visitors - Passport, J-1 Visa and I-94 Form
- Full Time Students - Passport, F-1 Visa, I-20 Form and I-94 Form
Copies of these documents may be either faxed or e-mailed to:
ERDC Security Office
ATTN: ORS HPC ACCESS
Dates of Visit: Start date - End date
Access to ORS Systems will expire 1 year from the start date, or on the expiration date of your DoD Contract, Permanent Residence Card or Visa, which ever comes earliest.
AWS GovCloud is a capability offered by the HPCMP strictly available for Open Research.
|Login Nodes||Compute Nodes|
|High Compute||Moderate Compute||GPU Accelerated|
|Total Cores | Nodes||2 | 1||4,104 | 114||4,096 | 512||4,096 | 128|
|Operating System||Amazon Linux 2|
|Core Type||AMD EPYC 7000||Intel Advanced Vector Extension 512||Intel Xeon E5-2686 v4|
|Core Speed||2.5 GHz||3.0 - 3.5 GHz||2.5 GHz|
|Memory/Node||16 GBytes||192 GBytes||42 GBytes||488 GBytes|
|Accessible Memory/Node||16 GBytes||192 GBytes||42 GBytes||488 GBytes|
|Memory Model||Shared on node.|
Frontera is a Dell EMC system located at the Texas Advanced Computing Center (TACC). It has 8,008 standard compute nodes, 16 large-memory compute nodes, and 90 GPU compute nodes (a total of 8,114 compute nodes or 450,784 compute cores). It has 1,594 TBytes of memory and is rated at 38.75 peak PFLOPS.
|Login Nodes||Compute Nodes|
|Standard Memory||Large Memory||GPU Accelerated|
|Total Cores | Nodes||224 | 4||448,448 | 8,008||1,792 | 16||704 | 90|
|Cores/Node||56||112||16 + 4 GPUs
(4 x 3,456 GPU cores)
|Core Type||Intel Xeon
Platinum 8280 Cascade Lake
+NVIDIA Quadro RTX 5000
|Core Speed||2.7 GHz||2.8 GHz|
|Memory/Node||192 GBytes||2.1 TBytes||128 GBytes
|Accessible Memory/Node||182 GBytes||2.0 TBytes||122
|Memory Model||Shared on node.||Shared on node.
Distributed across cluster.
|Interconnect Type||Ethernet / HDR InfiniBand|
Information on Frontera's production queues can be found at