IT platform teams’ responsibilities in 2024 are expanding radically. They’re being asked to provide cost-effective alternative solutions for the VMware virtualization technology that typically dominates enterprise IT. While they’re at it, could they please also stand up state-of-the-art AI platforms that can support generative AI in the data center and cloud that work as well as big cloud-managed AI services but for less money?

Those were the big questions hanging over some 6,000 attendees of Red Hat Summit in Denver May 6–9. Red Hat responded with a series of announcements of offerings — some immediate, some long-term — that are intended to answer them.

Red Hat led the conference, like every other tech vendor, with AI. The company has been in the AI space for years with its OpenShift Data Science offering, now rebranded as OpenShift AI, a Kubernetes-based AI platform used by Red Hat enterprise customers as the basis for bespoke platforms for data scientists. Red Hat’s parent company, IBM, markets essentially the same tech as watsonx.

Now, the IBM-Red Hat AI offerings are more deeply linked, as the new RHEL AI shows. Currently in developer preview, RHEL AI incorporates IBM’s Granite foundational model, enabling Red Hat to position OpenShift as a ready-made AI implementation for the enterprise and as an alternative to the click-and-go AI managed services from public cloud providers. Matt Hicks, CEO for Red Hat, also emphasized that the Granite models are being open-sourced by IBM Research — increasing competition for Meta’s Llama models. Importantly, Hicks pointed out indemnification options for businesses using those models for AI-enabled applications, putting more Red Hat (and IBM) skin in the enterprise AI adoption game.

But Red Hat’s AI offerings aim higher in the stack: Red Hat’s Podman AI Lab is poised to compete with the Ubuntu-Docker-NVIDIA combination that’s also aimed at winning developer hearts and minds. Red Hat also announced support for NVIDIA’s NIM microservices on OpenShift for generative AI use cases. It’s an open question whether developers accustomed to doing their own thing with Ubuntu will embrace RHEL AI and Podman AI Lab, especially given criticism of Red Hat’s decision to cut off RHEL source code from noncustomers. Red Hat is betting that they will, especially if the move is embraced — and perhaps enforced — by IT infrastructure teams. Even so, it could be a tough sell. Hicks and other Red Hat executives thus spent significant time talking about Linux and Red Hat’s perspectives on enterprise-grade operating systems.

While Red Hat will certainly get traction in cloud-based implementations of OpenShift, Red Hat is aiming for hybrid, enabling AI use cases on-premises — especially for customers facing spikes in cloud billing after turning on cloud managed services.

Red Hat Summit’s focus on the data center went beyond AI to more mundane matters, such as providing a path for customers seeking to mitigate VMware dependencies while keeping enterprise IT operating efficiently and affordably. Red Hat product leaders avoided overpromising — there’s no other virtualization stack with anything close to VMware’s capabilities today.

Adoption of the Red Hat OpenShift Virtualization platform to displace VMware, however, is a process of careful analysis, focused modernization, and surgical replacement, rather than rip-and-replace or lift-and-shift migration. The Red Hat customers showcased for switching from VMware to Red Hat OpenShift Virtualization were well resourced, tech-forward financial services giants. Most attendees would be advised not to try this at home until they can muster the necessary resources and skills and/or until product functionality matures. In the meantime, OpenShift, like other Kubernetes distributions, is increasingly adept at running virtual machines in other environments, providing a slower, less risky path to mitigating dependencies on VMware.

Other Red Hat Summit news didn’t quite splash in the tech press as much but grabbed the attention of platform teams and IT infrastructure bosses. These included new capabilities for the Ansible automation platform along with Lightspeed, a large language model-based knowledge management offering. While it got plenty of attention, Lightspeed is not scheduled for release until October 2024. This highlighting of coming attractions to showcase AI capabilities distracted from some of the here-and-now offerings unveiled at Summit. Another automation question hanging in the air was whether and how IBM’s proposed acquisition of HashiCorp will impact Ansible, given that Hashicorp’s Terraform and Ansible have both coexisted and competed for years.

Collectively, the slew of announcements and offerings keep Red Hat in contention as a leading enterprise IT platform alongside of (or hosted on) the hyperscale cloud providers, beyond the multicloud and hybrid Kubernetes platform that most of its immediate competitors have pursued. Even with the backing of IBM, competing at that scale is a very big bet, considering the huge investments of the hyperscalers in just AI alone. The announcements at Summit signal that Red Hat is taking on those challenges — but delivery on all fronts will be a major test.