GPU-Enabled Remote Workstation for VS Code

GPUs are graphic cards used to perform processing tasks. Using it increases the speed of training and the analysis of tasks that use artificial neural networks

GPU-enabled workstations are created for applications that require high levels of parallel processing. They include engineering, scientific or other applications. These can bring challenges to everyday work. When an organization gives GPUs enabled laptops, maintenance costs can be higher, and hardware can be underpowered for workloads. To make it easier for this professional, you can combine a GPU enabled in the cloud with the visual studio code remote development extension.

How to build a Cloud Workstation?

Visual Studio Code(VS Code) is gaining popularity due to its extensibility in single tap downloadable extensions. The remote Development extension (RDE)perfectly integrates a cloud-hosted development environment with all features set in the integrated development extension(IDE). When the extension is combined with Python extensions, data engineers and scientists can use a standard set of tools with a cloud-hosted GPU remote workstation from any device.

Oracle cloud infrastructure was invented to improve the performance that can be achieved with on-premises or private data options. It helps take away the maintenance burden and expenditure of hardware. Companies working in data science, artificial intelligence(AI), and other computing-intensive research can use a suite of tools. The tools allow designing and implementing a centralized cloud end-to-end workflow.

What does Oracle cloud GPU comprise of?

Oracle cloud GPU infrastructure is powered by NVIDIA's Tesla systems with Pascal(GPU2) and Volta(GPU3) architectures. For performant data access, files like BeeGFs, Spectrum Scale, and data solutions with automated databases are used. Additionally, instances are equipped with network interface cards. They can handle up to 25 GB/s when connected to high secure virtual cloud networks(VCNs) that are not virtual and shared.

After the team sets out the environment, they can save it as an easily deployable custom image. These images help save costs related to workstations' downtime and speed up the integration of new team members. Cloud workstations enable companies with remote employees access to the same performant infrastructure.

How to Deploy the Remote Workstation

Before deploying and connecting to the remote workstation, you require the VS Code and Remote Development extension. Then deploy a compute instance on Oracle Infrastructure from the VS Code Remote Workstation image available on the Oracle Cloud Marketplace. 

Deploy the instance in a subnet with public internet access that allows entry connections on port 22. The GPU-enabled VS Code workstation has all the standard packages and drivers needed to start. After the instance gets deployed, you can copy it to a public IP address. The reason is that you will need it to connect to the remote workstation. The only thing that remains is to follow the steps to finish the connection. Lastly, after connecting, you can explore attaching more volumes, for instance, specific storage that can be disconnected and reconnected to instances within an availability domain.

Cloud GPU instances and custom images can help achieve on-demand workstation solutions where datasets and access are standardized across the organization.


1 Blog posts