Skip to main content
Best Practices
Blueprints
Templates
Search
We found 3 articles tagged with "ai"
View All Tags
Go to Portfolio Navigator
Deploy the NVIDIA GPU Operator on CCE
The NVIDIA GPU Operator is a critical tool for effectively managing GPU resources in Kubernetes clusters. It serves as an abstraction layer over Kubernetes APIs, automating tasks such as dynamic provisioning, driver updates, resource allocation, and optimization for GPU-intensive workloads, thereby simplifying the deployment and management of GPU-accelerated applications. Its functionality extends to dynamic provisioning of GPUs on demand, managing driver updates, optimizing resource allocation for varied workloads, and integrating with monitoring tools for comprehensive insights into GPU usage and health. This guide outlines how to deploy the NVIDIA GPU Operator on CCE cluster. The process involves preparing GPU nodes, installing necessary components, configuring the cluster for GPU support, deploying an application leveraging GPUs, and verifying functionality.
Fuse-AI: e-Health from the Cloud
Fuse-AI is making medical diagnosis easier with artificial intelligence. A second opinion from the Open Telekom Cloud saves radiologists time, improves the quality of their diagnoses and reduces costs. The Hamburg-based entrepreneurs founded their start-up in 2015 and developed artificial intelligence that can detect indications of cancer – such as carcinomas – on MRI scan and assess whether a tumor is benign or malignant. And that doesn’t just save doctors time. “The biggest advantage is the improved quality of a diagnosis,” says Maximilian Waschka, one of the four Fuse-AI founders. “Our algorithm helps radiologists notice abnormalities on thousands of images more reliably.” The start-up estimates that its e-health solution can save health insurers at least 10 percent of the costs associated with MRI examinations.
Securely Expose Remote Ollama Endpoints to your Development Machine
Exposing Ollama endpoints directly from your cloud environment to your local development machine can be highly beneficial, especially when it comes to optimizing the use of expensive resources like GPUs and integrating them with local cost-effective development hardware.