🚧 GPU Runner - Bring Your Own GPU (BYOG)

📘

Please contact us to gain access to the Self-Hosted GPU Runner feature.

What is a GPU Runner?

GPU runners are systems that execute computational tasks using GPU (Graphics Processing Unit) resources that you have access to, whether they are physical hardware or cloud-hosted. This setup can be integrated into continuous integration and deployment (CI/CD) pipelines, enabling workflow execution within controlled environments.

Key Benefits

Cost Optimization

GPU runners allow organizations to maximize their GPU investments, whether through owned hardware or dedicated cloud resources. This approach can lead to significant cost savings, especially for organizations with consistent GPU workloads.

Enhanced Security
Running workflows on controlled GPU infrastructure provides better security for sensitive operations. Users maintain full control over their execution environment, ensuring code, data, and proprietary algorithms remain protected.

Customization Freedom
Self-hosted GPU runners offer complete control over the environment configuration, including:

• Software stacks and dependencies
• Custom library versions
• System-level optimizations
• Resource allocation policies

Flexibility
The GPU resources can be:

• Physical hardware in your datacenter
• Cloud-hosted GPU instances you manage
• Hybrid setups combining both approaches

This flexibility allows organizations to choose the most suitable infrastructure strategy while maintaining control over their compute resources.

Try It Out

Check out how to set up your environment to install runners and train models on your own hardware.