How Red Hat partners are powering the next wave of enterprise AI

The pace of AI innovation is accelerating, and with the launch of Red Hat AI 3, we're reminded that turning this potential into enterprise reality requires a robust, open ecosystem built on choice and collaboration. Our goal has always been to provide a consistent, powerful platform for AI that works with any model, on any accelerator, and across the hybrid cloud. Today, we're thrilled to highlight the momentum from our partners, who are working alongside us to build out the future of open, hybrid AI on Red Hat.

The Red Hat partner ecosystem is the engine that will deliver the generative AI (gen AI) and agentic capabilities that customers need for broad market adoption. It’s about bringing together the best in hardware, software, and services to create a whole that is far greater than the sum of its parts.

Powering fast, flexible and efficient inference 

The launch of Red Hat AI 3 is centered on driving enterprise AI inference, expanding model choice and enabling open models for optimized cost and flexibility – so organizations can go from training to “doing.” And Red Hat partners play a critical role in making this happen. 

To deliver AI inference at scale using vLLM and Kubernetes, the llm-d open source project is now generally available as part of Red Hat OpenShift AI 3.0, powered by a coalition of leading gen AI model providers, AI accelerators and premier AI cloud platforms. Founding contributors include CoreWeave, Google Cloud, IBM Research, and NVIDIA, with additional support from partners like AMD, Cisco, Hugging Face, Intel, Lambda, and Mistral AI. Since the project’s introduction earlier this year, Microsoft, Oracle and WEKA have also become active contributing members. 

As Large Language Models (LLMs) become the foundation for a wide range of gen AI applications, Red Hat is introducing the Partner Model Validation Guide to empower our partners and provide greater choice to customers. This guide outlines a standardized, step-by-step process for Red Hat partners to benchmark their LLMs for inclusion in the Red Hat OpenShift AI model catalog.

Leveraging the Partner Model Validation Guide, model provider partners can generate the data needed for validation using a prescribed open source toolchain, including vLLM for serving, GuideLLM for performance benchmarking, and LM Evaluation Harness for accuracy evaluations. Once submitted and validated by our team, these partner models, complete with their performance and quality metrics, will be made visible and available to all Red Hat OpenShift AI customers. This initiative enriches our model catalog, offering customers more transparent insights and guidance at the model level, a broader selection of best-in-class open source models to build their AI applications, and market visibility for our model provider partners.

Accelerating agentic AI deployments

Ultimately, the goal is to get AI applications into production where they can deliver business value. Red Hat AI, combined with our partner ecosystem, provides the essential capabilities for developing, deploying, and managing agentic AI. This is where our systems integrators and services partners play a crucial role. Red Hat is engaged with global leaders like Accenture, HCL, Kyndryl, IBM Consulting, Infosys and NTT DATA to support customers in scaling these applications and optimizing inference to run AI everywhere – from the datacenter to the edge. 

With a solid foundation and enabled infrastructure, the focus shifts to the tools and models that power intelligent applications. An AI strategy is only as good as the ecosystem of data platforms, vector databases, security tools, and developer utilities that support it. Red Hat’s partner landscape is rich and diverse, providing the essential components for building modern AI workflows.

Model Context Protocol (MCP)

With the launch of Red Hat AI 3, we've curated a collection of MCP servers. This enables ISVs to connect their tools and services directly to Red Hat AI. Partners like CyberArk, Dynatrace, Elastic, EnterpriseDB, Palo Alto Networks, and others are working with Red Hat to showcase their MCP servers, helping customers build sophisticated agentic applications with trusted tools. 

Llama Stack

Red Hat’s work with Meta’s open source Llama Stack is foundational to empowering developers with Red Hat AI 3. By actively contributing to this open source project, Red Hat and its partners are helping to enhance a complete set of standardized tools that revolutionize the gen AI application lifecycle. This helps ensure developers have a compelling, robust, and adaptable environment to accelerate the creation and deployment of innovative AI solutions, giving our partners and customers a clear path to production. By reducing the complexity of integrating various AI services and tools, Llama Stack enables developers to focus on innovation rather than infrastructure challenges.

Scaling AI across the hybrid cloud

AI Accelerators  

Options are important to cover the widest range of capabilities, which is why we've deepened our collaborations with accelerator partners like AMD, Google Cloud, and Intel. 

We’re working with NVIDIA to integrate NVIDIA’s performance-optimized AI computing, networking, and software with Red Hat AI, helping customers to build, deploy, and manage AI at scale and optimize infrastructure across diverse workloads. Red Hat and NVIDIA continue to drive the expansion of AI infrastructure with the support for NVIDIA GB200 NVL72 infrastructure, targeting high-end AI training and inference at scale. Furthermore, the NVIDIA RTX PRO(R) 6000 Blackwell Server Edition GPUs support Red Hat AI, Red Hat OpenShift, and Red Hat OpenShift Virtualization, accelerating a range of enterprise workloads, including enterprise AI and VDI graphics, with air-cooled GPUs and servers. 

Our collaboration with AMD is powering more efficient gen AI by enabling AMD’s portfolio of x86-based processors and GPUs on Red Hat AI, driven by our activations in upstream communities like vLLM and llm-d to deliver enhanced performance and GPU support. 

In addition, Red Hat continues to build upon its longstanding collaboration with Intel to support Red Hat AI across its expanding hardware platform and accelerator portfolio, including Intel® Gaudi® AI accelerators. 

This commitment to choice extends across the hybrid cloud. In this area, we are actively working with Google Cloud to bolster TPU support on Red Hat AI, extending our shared vision for community-powered innovation through projects like llm-d and vLLM.

Infrastructure Solutions 

But accelerators are just one piece of the puzzle. Customers need validated, integrated solutions. That's where our OEM server partners come in, transforming the theoretical potential of AI into tangible enterprise reality. While the foundation of AI lies in accelerators on the bottom and language models on the top, the practical deployment and management of AI at scale requires robust, integrated solutions. 

Red Hat is actively collaborating with industry leaders, including Cisco, Dell Technologies, HPE, and Lenovo, to simplify the deployment and management of AI across hybrid cloud environments. With Dell, we are working to help ensure that their powerful server and storage solutions are optimized for Red Hat AI, providing customers with reliable and scalable infrastructure for their AI workloads. This includes joint efforts to enable Red Hat OpenShift and Red Hat OpenShift AI for the Dell AI Factory with NVIDIA, which brings together optimized hardware and software capabilities to drive more seamless AI deployments that can handle the most demanding computational tasks.

Similarly, our collaboration with Cisco combines their strengths in networking, security, compute, and observability with Red Hat OpenShift as the application platform for containerized AI workloads within Cisco AI PODs, the core AI infrastructure building block of Cisco Secure AI Factory with NVIDIA. Red Hat AI builds on this foundation to deliver a consistent, automated environment for AI model development, training, and inferencing. The result is a full-stack AI platform that is scalable, high-performing and secure.

We are expanding AI infrastructure reference designs and offerings across our entire infrastructure ecosystem. These collaborations are essential in providing customers with the comprehensive solutions needed to operationalize AI, helping ensure that the necessary hardware, software, and services can be integrated to accelerate AI adoption and innovation.

Our vast ecosystem of global distributors and value added resellers (VARs) is instrumental in building repeatable solutions at scale for the enterprise. Partners like Ahead, Arrow, TD Synnex, WWT, and many others are aligning to Red Hat AI within their solution portfolio.

Delivering a leading AI ecosystem

Through strategic collaborations with partners across hardware, software, and services, Red Hat is providing a consistent, powerful platform for AI that offers choice and flexibility across the hybrid cloud. This collaborative approach, from optimizing distributed inference with llm-d to validating language models with the Partner Model Validation Guide and accelerating agentic AI deployments, helps ensure that customers can operationalize AI at scale and unlock its full business value.