Cloudera launches AI Inference, a new service that leverages Nvidia technology for faster processing of AI models. The service focuses on securely managing and deploying large-scale AI applications, including Generative AI (GenAI), and offers up to 36 times faster performance.
Cloudera AI Inference is one of the first AI inference services to use Nvidia NIM microservices. This integration, part of the Nvidia AI Enterprise platform, enables faster deployment and management of large-scale AI models. It allows organizations to get GenAI out of the pilot phase and into real-world applications more efficiently. The service helps developers build and manage Large Language Models (LLMs) with advanced security and scalability.
The collaboration between Cloudera and Nvidia offers improved performance through the use of Nvidia’s Tensor Core GPUs. This results in 36 times faster processing than traditional methods. The new service offers direct integration of the user interface and APIs with Nvidia NIM microservice containers. This reduces the need for complex tools such as CLIs, simplifying the management and monitoring of AI models.
Security and scalability central
A key feature of Cloudera AI Inference is its emphasis on security and privacy. The service prevents sensitive data from leaking into vendor-hosted AI modeling services by giving companies control over the development and deployment of their own AI models. In addition, the service supports both on-premises and cloud-based deployments, providing flexibility for organizations requiring strict regulatory compliance.
The service is equipped with features for scalability, monitoring and security. This allows organizations to efficiently deploy AI models while meeting compliance standards and governance requirements. Automatic scalability and real-time performance tracking help detect and resolve problems quickly, ensuring optimal resource management. Cloudera’s integration with Nvidia provides a solution for companies looking to deploy reliable AI without complex do-it-yourself approaches.