Red Hat has made huge strides in virtualization. Managing virtual machines (VMs), containers and, more recently, heavier AI workloads demands constant attention from IT. At the same time, the virtualization issue is starting to extend beyond the traditional IT department because of the rising costs tied to virtualization challenges. How can you best respond to this? We discussed it with Mike Barrett, Vice President and General Manager of Red Hat Hybrid Platforms.
The problem many infrastructure teams face today goes significantly further than the widely discussed rise in license costs. IT departments currently have to keep three fundamentally different types of workloads up and running. They are balancing classic VMs for essential legacy applications, fast containers for modern cloud-native development, and heavy GPU-intensive workloads for artificial intelligence. On traditional platforms this requires three separate operational models, each with their own management teams, security protocols and specialized tooling. According to Red Hat, this fragmented approach is unsustainable in the long term and, above all, extremely inefficient.
Red Hat is therefore sharply focused on consolidation. OpenShift is meant to bring these three worlds together on a single platform, driven by one universal operational model and a consistent security approach. VMs, containers and advanced AI applications run side by side without issues and are managed and secured through exactly the same interface. As a result, the complexity for administrators is drastically reduced.
From IT headache to boardroom business case
That virtualization is a high priority on the IT side is not surprising in itself. However, it has also become a topic of discussion outside the IT department. This is mainly due to the far-reaching changes in VMware’s licensing models following Broadcom’s acquisition at the end of 2023. The new licensing structures at VMware have put the entire market in motion. Red Hat responded proactively by enticing VMware customers with migration options, in some cases delivering cost savings of up to 70 percent.
According to Barrett, the urge to migrate now does not primarily stem from a deeply rooted technical ambition but from a simple, unavoidable economic calculation. Organizations can no longer justify the exponentially increased costs of their current virtualization solution to the board. Yet the impact of this market shift is not felt everywhere at the same pace; it is moving through the sector in waves. Barrett points out that many large enterprise organizations work with long-term contracts of five to eight years. This means the financial shockwave is only now reaching some companies. As a result, many organizations currently find themselves in a transition period, which explains why Red Hat is still attracting new customers that are actively seeking strategic support in redesigning their IT landscape.
What has fundamentally changed is where the final decision on virtualization is made. Previously, it was a purely technical and fairly invisible issue for the IT department, buried deep in the data center. Now it is prominently on the agenda in the boardroom. “If you have a business case with real numbers, it becomes much easier to introduce a larger platform that the entire company can agree on,” says Barrett. The financial argument has become compelling. It makes the business decision far more attractive than a simple one-to-one replacement of the underlying hypervisors for purely technical reasons.
Rethinking the cloud strategy
When undertaking a large-scale infrastructure migration, the public cloud was often seen as a logical step. This raises the question of why organizations do not simply move all their workloads to AWS, Azure or Google Cloud. Red Hat presents three key arguments that convince many large enterprises to opt for a hybrid model instead.
First, there is the issue of operational consistency. Organizations that build their IT architecture fully on the specific, proprietary services of a single hyperscaler will inevitably face vendor lock-in and fragmented operations once they later decide to expand into another environment or provider. Each individual cloud uses a completely different operational model, different tooling and other network primitives. In contrast, Red Hat OpenShift acts as a universal abstraction layer that simply works in the same way everywhere. Whether the platform runs on private infrastructure in the company’s own data center, in a public cloud, at the network edge or in fully isolated environments, the management experience remains exactly the same.
A second, closely related argument is the drive for independence and data sovereignty. Organizations that standardize deeply on the proprietary services of a major tech vendor inevitably give up negotiating power and control. An open hybrid cloud model is meant to offer an alternative to this dependence. Barrett notes that this topic is now a major concern in boardrooms. In independent evaluations where platforms are tested across multiple infrastructures, Red Hat’s technology consistently scores highly in this area. The terminology used proves crucial: where a dry technical story about hybrid clouds sometimes fails to resonate, the very same story almost always finds support when framed in the current context of digital sovereignty.
Finally, hard law and regulation play an undeniable role. For organizations in heavily regulated sectors such as financial services, healthcare, critical government functions and telecom, a fully public cloud strategy is often impossible because of strict national and international compliance requirements. For these specific sectors, a well-designed hybrid cloud is a workable, legal, and secure IT strategy they can adopt.
Migration at scale demands smarter tooling
Migrating an entire virtualization environment, however, is a project that takes time. In practice, it often affects hundreds of business-critical applications and brings real operational risks with it. Engineers and administrators face the challenge of keeping the legacy environment running with full availability while simultaneously designing, building and configuring the new architecture. All this usually has to happen without any structural expansion of the workforce and under the condition that not a single Service Level Agreement may be violated.
To effectively rein in this complexity, Red Hat developed the Virtualization Migration Assessment. Using this in-depth service, Red Hat has already analyzed more than one million VMs worldwide, resulting in a wealth of raw data and invaluable real-world experience. The software gives organizations complete, transparent visibility into their entire application landscape before a single workload is physically moved. The tool connects directly to the existing virtualization environment and assesses every VM carefully and individually. The system can almost immediately recognize whether a particular VM uses a complex clustered file system, which requires significantly more migration time and preparation, or whether it simply runs on default settings and can therefore be moved automatically in a batch. For each VM, the software delivers a concrete, reliable estimate of the required migration time. This enables the IT department to set data-driven priorities, create phased migration plans with fallback options and allocate budgets.
Catching up on the work floor
Although the Linux KVM hypervisor under OpenShift Virtualization is the most deployed hypervisor technology in the world, large organizations did express a number of additional wishes during the migration wave in and after 2023. Virtualization administrators wanted a familiar left-hand navigation bar, clear features such as multi-selection and effective buttons for pausing, starting and stopping systems. Red Hat realized the platform had to adapt to the user’s way of working, and absolutely not the other way around. This ClickOps style of working was therefore developed at record speed and fully integrated into the standard OpenShift interface.
At the same time as the interface adjustments, the underlying network stack also had to be significantly expanded to be truly enterprise-ready. Specialized virtualization teams want every tenant to have their own isolated subnet, full support for hard-coded IP addresses and the ability to migrate a complex network topology one to one to the new platform. This wish list triggered a strong chain reaction of complex technical requirements in the backend, including extensive integrations for VLAN trunking, dynamic BGP routing, EVPN and granular load balancing. Red Hat OpenShift took that step by introducing user-defined networks for VMs, which immediately eliminated a large portion of these day-to-day pain points.
Ready for live migrations
A more recent major development took place at the end of 2025. That is when the much requested capability for live migration of VMs between two completely different Kubernetes clusters became available, without noticeable downtime for the user. Efficient differential backups were also added to the core, with the system storing only the raw data that has actually changed. This functionality integrates with platforms such as Veeam, Commvault and Rubrik.
In addition, version 4.21 introduced advanced Dynamic Resource Allocation for GPUs, which definitively prepares and scales the platform for the extremely heavy AI workloads of the near future. The next version of OpenShift will add memory overcommitment. As a result, the underlying Kubernetes engine learns to treat memory allocation for VMs fundamentally differently than for standard lightweight containers. Combined with right-sizing dashboards, which give administrators immediate visual insight into where expensive resources may be wasted unnecessarily, Red Hat aims to provide an answer to the hardware scarcity in the market.
To lower the entry barrier for traditional IT environments, Red Hat also launched Red Hat OpenShift Virtualization Engine in 2025. This is a special, streamlined edition of OpenShift focused on running VMs in a stable way, without forcing administrators to immediately deal with the broader container layer. It serves as a stepping stone for companies that want to modernize but still find the jump to full containerization or AI adoption too big.
Milestones
That the broad strategy of offering a single uniform platform for VMs, containers and AI workloads is truly successful is evident from the hard numbers the company recently shared. Red Hat proudly announced an important milestone: OpenShift has passed the symbolic threshold of $2 billion in Annual Recurring Revenue. In addition, there has been a 400 percent increase in migrated virtual machines and more than 70 percent growth in the number of new paying customers. This shows that the desire to modernize virtualization remains very much alive today.
Also read: Chris Wright: Metal-to-agent is the foundation of scalable enterprise AI