Development Environment

For building and testing applications for Fleet Command, developers should build a development environment that matches the NVIDIA Fleet Command environment.

Fleet Command supports NVIDIA-Certified Systems with these programs. NVIDIA works with leading OEMs to validate that their systems provide an expected level of performance for AI workloads and that they can be secured and remotely managed for edge-use cases.

System requirements vary from application to application, but the minimum standards for the development environment are:

• Single socket x86-based systems with six cores

• One NVIDIA T4

• 10GbE NIC

• 48GB RAM

More information about configuring systems is available in the NVIDIA-Certified Configuration Guide .

While other systems and GPUs can be used for a development environment, your options for setting up the software environment may be limited. End-to-end testing may not be possible with Fleet Command without a supported systems.

Development systems need access to the internet to download the components necessary to set up the development environment.

Fleet Command Stack is a secure and maintained software stack based on NVIDIA Cloud Native Core.

To develop applications compatible with Fleet Command, we recommend using the version of Cloud Native Core that correlates with the specific version of Fleet Command location that you will be deploying to.

For a mapping of Fleet Command Stack versions to Cloud Native Core versions, refer to the following table:

Applications developed on a specific version of Cloud Native Core or Fleet Command Stack may require changes to run on other versions. For more information, refer to the System Updates section.

For information on how to set up Cloud Native Core, the install guides can be found here: https://github.com/NVIDIA/cloud-native-core/tree/master/install-guides