GPU Worker Support Now Available in Kubernetes Service
17 September, 2025
Catalyst Cloud has released significant updates to the Catalyst Cloud Kubernetes Service (CCKS) that deliver seamless GPU acceleration capabilities for cloud-native workloads. These improvements remove traditional technical barriers and streamline the deployment of GPU-powered applications within Kubernetes environments.
Key Improvements
The latest CCKS versions include several important enhancements:
Automated Driver Management: The service now features automatic NVIDIA driver detection and installation during node boot processes, eliminating manual configuration requirements.
Enhanced Compatibility: Full compatibility with NVIDIA's gpu-operator ensures seamless integration with existing GPU workflows and tooling.
Quality Assurance: Built-in testing within publishing pipelines prevents regressions and maintains service reliability.
Version Availability: GPU acceleration is available in patch versions v1.31.12, v1.32.8, and v1.33.4 and later.
Addressing Technical Challenges
Previously, users faced significant complexity when selecting and installing NVIDIA drivers on Flatcar OS, which powers the Kubernetes service. This process was often technically challenging or impossible for many organisations. The enhanced CCKS versions now automatically handle this entire process before nodes join the Kubernetes cluster, removing these implementation barriers.
Business Impact
This development enables organisations to leverage GPU compute power for artificial intelligence, machine learning workloads, data processing, and high-performance computing applications directly within their Kubernetes clusters. The automation eliminates traditional setup complexities, allowing teams to focus on application development rather than infrastructure management.
The fully managed CCKS platform continues to provide 24/7 monitoring and management, enabling organisations to maximise productivity while Catalyst Cloud handles the underlying infrastructure complexity.
For detailed implementation guidance, visit the GPU acceleration documentation or learn more about the Catalyst Cloud Kubernetes Service.