Red Hat updates AI portfolio to enhance hybrid cloud performance

Red Hat has updated its Artificial Intelligence (AI) portfolio, enhancing Red Hat OpenShift AI and Red Hat Enterprise Linux AI. These updates target the optimization of AI model deployment and performance within hybrid cloud environments, addressing challenges related to cost reduction and data integration for enterprises.

The latest version of Red Hat OpenShift AI introduces several features, including distributed serving through the vLLM inference server, which allows IT teams to utilize multiple GPUs for efficient model serving. It also offers an end-to-end model tuning experience using InstructLab, enhancing the scalability and manageability of AI models.

AI Guardrails, a technology preview included in the recent release, aims to improve Large Language Model (LLM) accuracy and performance by detecting potentially harmful user interactions and model outputs. Additionally, a new model evaluation component supports data scientists in benchmarking the performance of their AI models across various tasks.

Further, Red Hat Enterprise Linux AI (RHEL AI) serves as a foundation model platform aimed at consistently developing and deploying LLMs across hybrid clouds. The recent RHEL AI 1.4 update introduces Granite LLMs with multilingual support and new contributions interfaces to simplify data ingestion.

As part of its commitment to educating organizations about AI, Red Hat is now offering no-cost AI Foundations online training courses for users at all levels. This initiative is designed to foster a better understanding of how AI can enhance business operations.

Red Hat OpenShift AI 2.18 and RHEL AI 1.4 are now generally available. Enterprises interested in utilizing the updated AI solutions can find additional details in Red Hat’s official documentation.