Red Hat Expands AI Capabilities with RHEL AI 1.3 Release

Red Hat Unveils RHEL AI 1.3 with Enhanced AI Model Support

Red Hat, an IBM subsidiary, has released version 1.3 of Red Hat Enterprise Linux AI (RHEL AI), a specialized Linux platform designed for developing, testing, and running artificial intelligence large language models (LLMs). This latest release significantly expands AI model support and introduces new features to streamline enterprise AI development and deployment 1

1

2

Key Enhancements in RHEL AI 1.3

The new version brings several notable improvements:

1. Granite LLM Support: RHEL AI 1.3 now supports Granite 3.0 8b, an 8 billion-parameter model capable of handling multiple languages and offering code generation and function-calling capabilities 1

   1

   .

2. Improved Data Preparation: The release incorporates Docling, an open-source project from IBM Research, which enables the conversion of various document formats (PDFs, manuals, slide decks) into specialized data formats like JSON and Markdown 1

   1

   2

   2

   .

3. Expanded Deployment Options: RHEL AI 1.3 offers enhanced choices for hybrid cloud deployment, allowing for more flexible implementation across diverse environments 1

   1

   .

4. Hardware Acceleration: The platform now includes support for Intel's Gaudi 3 AI accelerator, complementing existing support for NVIDIA and AMD GPUs 1

   1

   2

   2

   .

Advanced Features and Capabilities

RHEL AI 1.3 introduces several advanced features:

• Context-aware Chunking: This natural language processing method breaks down text into smaller, meaningful segments while preserving context, leading to more coherent and contextually appropriate AI responses 1

  1

  .

• Parallelized Serving: Red Hat OpenShift AI, which natively supports RHEL AI, now enables parallelized serving across multiple nodes with vLLM runtimes, allowing for real-time handling of multiple requests 1

  1

  .

• Dynamic LLM Parameter Adjustment: Users can now dynamically alter an LLM's parameters during serving, including sharding the model across multiple GPUs or quantizing it to a smaller footprint 1

  1

  .

Related Stories

Red Hat Launches Enterprise Linux AI, Partners with Dell for Open-Source AI Workloads

Red Hat Unveils AI 3: A Leap Forward in Enterprise AI Deployment and Management

Red Hat launches AI Enterprise platform to accelerate hybrid cloud AI deployments at scale

Red Hat Acquires Neural Magic to Enhance AI Performance Across Hybrid Cloud Environments

Implications for Enterprise AI Adoption

Red Hat's focus on supporting smaller, open-source-licensed models aligns with the growing trend of enterprise AI adoption. A recent International Data Corp. report found that 61% of enterprises plan to use open-source foundation models for generative AI 1

1

Joe Fernandes, Red Hat's AI business unit vice president and general manager, emphasized the role of service partners and systems integrators in helping companies explore and integrate AI use cases cost-effectively. He stated, "AI in general and generative AI in particular is complicated, so our services partners and systems integrators are the ones helping companies explore and integrate different use cases in a cost-effective way" 2

2

Future Developments

Red Hat has outlined plans for future RHEL AI releases, including:

1. Full support for Granite 3.0 8b's non-English language, code generation, and function calling capabilities 2

   2

   .
2. Additional document format support for Docling 2

   2

   .
3. Integration for retrieval-augmented generation pipelines 2

   2

   .
4. InstructLab knowledge tuning enhancements 2

   2

   .

As AI continues to evolve rapidly, Red Hat's RHEL AI 1.3 represents a significant step forward in providing enterprises with the tools and flexibility needed to develop and deploy AI solutions effectively across hybrid cloud environments.