The new AI stack: Choice, control, and production-ready innovation

In the next decade, AI will redraw the map of technology ecosystems. As we traverse what Forrester is calling the "seventh wave" of major technological change—driven by generative and agentic AI—C-suite executives are facing a daunting transition. The difference between falling behind and harnessing this wave of change is your strategy for the AI computing stack.

At Red Hat, our mission remains centered on open source principles: Collaboration, transparency, and choice. We believe that for AI to truly deliver on its promise of productivity and business value, it cannot remain a proprietary "black box." It must be grounded in the same solid engineering principles that made Linux, Kubernetes, and other open source innovations the foundation of the modern enterprise.

Decoding the new ecosystem: Insights from Forrester

To help make sense of these changes, I highly recommend reading, and understanding the implications of, the new Forrester report, AI Powers A New Computing Ecosystem. It provides an essential roadmap for this transition. The report highlights how AI chips, models, and agents are dismantling the legacy cloud and software stacks we’ve relied on for years.

Forrester identifies 8 distinct categories of providers—from chipmakers to agentic platforms—that are now colliding as they race to shape the future of computing. Understanding these shifting structures, and the pressures and incentives that are emerging, is critical for any leader looking to maximize the value they create with their partners rather than getting distracted by the noise and hype.

This requires a fundamental shift in how we approach the AI stack. As the ecosystem expands, the challenge for a CIO or CTO is no longer just about choosing a cloud provider. It is about architecting a stable, scalable, and controllable stack across 5 critical layers: Experience, orchestration, data, intelligence, and infrastructure. Getting each of these layers right helps provide the foundation needed for resiliency and adaptability.

Why Red Hat? Any model, any accelerator, any cloud

The market is currently dominated by proprietary providers, but history tells us that open solutions ultimately prevail. Red Hat is uniquely situated because we offer a neutral, hybrid operating layer. Our strategy is built on choice, flexibility, and control:

1. Any model: Whether you are using IBM’s Granite models, Llama, Mistral, models from other major providers, or specialized internal models, platforms like Red Hat AI provide the flexibility to build, tune, and deploy them backed by Red Hat’s Open Source Assurance program for intellectual property assurance.
2. Any accelerator: We believe in "silicon neutrality." We work closely with the entire ecosystem—including NVIDIA, AMD, and Intel—to help you maintain choice in your hardware path.
3. Any cloud: Your AI strategy must follow your data. Whether that data lives on-premises for "sovereign AI" requirements or across multiple public clouds, Red Hat provides a consistent operational model.

The power of collaboration: Red Hat and NVIDIA

Red Hat’s close collaboration with NVIDIA is a case study in how this can work. A critical component of this new infrastructure is the hardware and software substrate that powers it. This is why our expanded work with NVIDIA is so transformative. We are pairing enterprise open source with NVIDIA’s rack-scale AI advances to deliver faster, production-ready innovation.

By providing "Day 0" support for the NVIDIA Vera Rubin platform, we are making sure that the world’s leading enterprise Linux platform, Red Hat Enterprise Linux (RHEL), is optimized for the next generation of AI supercomputing as soon as it’s available. This means providing a hardened, consistent foundation with integrated lifecycle management so teams can focus on innovation rather than infrastructure friction.

Moving from proof of concept to production

According to Gartner®, "At least 30% of generative AI (GenAI) projects will be abandoned after proof of concept by the end of 2025, due to poor data quality, inadequate risk controls, escalating costs or unclear business value." To avoid this "POC purgatory," executives must prioritize a platform approach that addresses technical debt and simplifies the path to production AI.

The first step is modernizing your foundation. As you explore the insights in the Forrester report, remember that an enterprise built to adapt to disruption is an enterprise built for AI. Red Hat is here to be your trusted partner in approaching these high-stakes decisions, helping you to keep your AI strategy as durable as the business you are building. Together, we can manage the complexity and shifting boundaries as AI innovation continues to unfold. Co-innovation is the way forward.

Want guidance on other tech challenges?

Discover more expert perspectives in The Executive’s Collection—where Red Hat leaders and industry analysts help you make smarter decisions and stay ahead of disruption. Read the collection.

Gartner Press Release, Gartner Predicts 30% of Generative AI Projects Will Be Abandoned After Proof of Concept by End of 2025, July 29, 2024 
GARTNER is a trademark of Gartner, Inc. and/or its affiliates.