Attention A T users. To access the menus on this page please perform the following steps. 1. Please switch auto forms mode to off. 2. Hit enter to expand a main menu option (Health, Benefits, etc). 3. To enter and activate the submenu links, hit the down arrow. You will now be able to tab or arrow up or down through the submenu options to access/activate the submenu links.

VA Technical Reference Model v 21.9

CUDA Deep Neural Network (cuDNN)
CUDA Deep Neural Network (cuDNN) Technology

General InformationGeneral Information help

Technologies must be operated and maintained in accordance with Federal and Department security and privacy policies and guidelines. More information on the proper use of the TRM can be found on the TRM Proper Use Tab/Section.

Website: Go to site
Description: The Compute Unified Device Architecture (CUDA) Deep Neural Network (cuDNN) is a graphics processing unit (GPU) accelerated library of primitives for deep neural networks. cuDNN provides highly tuned implementations for standard routines. cuDNN features customizable data layouts, supporting flexible dimension ordering, striding, and subregions for the four-dimensional tensors used as inputs and outputs to all of its routines. This flexibility allows easy integration into any neural network implementation and avoids the input/output transposition steps sometimes necessary with GEneral Matrix to Matrix Multiplication (GEMM) based convolutions. cuDNN is part of the NVIDIA Deep Learning Software Development Kit (SDK).

This technology uses a proprietary database.

The TRM decisions in this entry only apply to technologies and versions owned, operated, managed, patched, and version-controlled by VA. This includes technologies deployed as software installations on VMs within VA-controlled cloud environments (e.g. VA Enterprise Cloud (VAEC)). Cloud services provided by the VAEC, which are listed in the VAEC Service Catalog, and those controlled and managed by an external Cloud Service Provider (i.e. SaaS) are not in the purview of the TRM. For more information on the use of cloud services and cloud-based products within VA, including VA private clouds, please see the Enterprise Cloud Solutions Office (ECSO) Portal at: https://dvagov.sharepoint.com/sites/OITEPMOECSO
Technology/Standard Usage Requirements: Users must ensure their use of this technology/standard is consistent with VA policies and standards, including, but not limited to, VA Handbooks 6102 and 6500; VA Directives 6004, 6513, and 6517; and National Institute of Standards and Technology (NIST) standards, including Federal Information Processing Standards (FIPS). Users must ensure sensitive data is properly protected in compliance with all VA regulations. Prior to use of this technology, users should check with their supervisor, Information Security Officer (ISO), Facility Chief Information Officer (CIO), or local Office of Information and Technology (OI&T) representative to ensure that all actions are consistent with current VA policies and procedures prior to implementation.
Section 508 Information: This technology has not been assessed by the Section 508 Office. The Implementer of this technology has the responsibility to ensure the version deployed is 508-compliant. Section 508 compliance may be reviewed by the Section 508 Office and appropriate remedial action required if necessary. For additional information or assistance regarding Section 508, please contact the Section 508 Office at Section508@va.gov.
Decision: View Decisions

Decision Source: TRM Mgmt Group
Decision Process: One-VA TRM v21.3
Decision Date: 03/02/2021
Introduced By: TRM Request
Vendor Name: NVIDIA
- The information contained on this page is accurate as of the Decision Date (03/02/2021).