Red Hat launches AI 3.5 to scale enterprise AI workloads

Red Hat has announced the release of Red Hat AI 3.5, a platform update designed to transition enterprise AI from experimental pilots to production-ready, mission-critical infrastructure. The update focuses on providing operational rigor through enhanced safety, observability, and multi-tenancy controls.

Key features include EvalHub, which allows organizations to benchmark models for safety and regulatory compliance before deployment. The platform's AI Catalogue has been expanded with over 20 validated models from providers such as Google, NVIDIA, and Alibaba Cloud, including tags for tool-calling capabilities.

To address resource management, the release introduces new observability dashboards for monitoring inference health, GPU utilization, and model performance. It also features advanced multi-tenancy capabilities, such as fair-share GPU scheduling and priority-aware serving. These tools allow enterprises to manage shared GPU infrastructure by prioritizing mission-critical workloads over background tasks and providing hardware-to-software isolation for sensitive data workloads.