Your infrastructure budget is about to get weird.
Not because technology is changing. Technology always changes. But because the economic forces that determine where you run workloads, which platforms you standardize on, and how you pay for compute are all shifting at once.
Cloud pricing that made sense in 2023 looks different after two years of hyperscaler price increases. VMware licensing that was predictable for a decade became unpredictable overnight. Hardware costs that used to follow Moore's Law now follow geopolitics, tariff schedules, and RAM market volatility.
The infrastructure decisions you make in 2026 won't just be about technical capability. They'll be about economic resilience. And the enterprises that treat this as purely a technology conversation are going to discover they're having the wrong conversation.
We asked five e360 experts who architect, implement, and operate enterprise infrastructure what they're seeing. Their predictions aren't about which vendor has the best roadmap. They're about which economic realities you can't ignore.
A Field Chief Technology Officer at e360, sees the clearest trend emerging: "I believe data center repatriation will become more prevalent across the enterprise and commercial segments to allow for better economic benefits while modernizing and improving operational efficiency. This enables basic cloud-like functionality without stifling innovation and velocity."
Let's be clear about what this is and what it isn't. This isn't an anti-cloud backlash. This isn't a rejection of cloud architecture principles. This isn't nostalgia for the pre-cloud era.
This is math. When cloud costs are rising, when data gravity makes egress fees prohibitive, when your workload profiles don't match the pricing models that hyperscalers optimize for, bringing workloads back to modernized data centers isn't ideological. It's economic.
The key phrase in Hen's prediction is "cloud-like functionality." Organizations aren't repatriating to go back to legacy data center operations. They're building modern, software-defined infrastructure that delivers cloud-like agility, automation, and elasticity without cloud-like bills.
The winning strategy in 2026 isn't "cloud" or "on-prem." It's "cloud where it makes economic sense, modernized data center where it doesn't, and the operational discipline to know the difference."
Sean Ibarra, Account Executive at e360, hears the same three questions from clients constantly: "VMware alternatives, Terraform and IBM questions and alternatives, and 'we have AI...now what?'"
These aren't unrelated questions. They're all symptoms of the same problem: platform lock-in meeting economic pressure.
The VMware Question: Broadcom's acquisition of VMware changed the licensing model overnight. Organizations that built their entire virtualization strategy around predictable VMware costs now face renewal negotiations that look nothing like previous cycles. The question isn't whether VMware is still technically capable. The question is whether the new economic model is sustainable for your environment.
VMware alternatives exist. OpenStack, Proxmox, open-source KVM platforms, and hyperscaler-native virtualization all work. But migrating off VMware isn't a weekend project. It's a multi-quarter initiative that touches every workload, every operational runbook, every backup and disaster recovery process, every monitoring integration.
2026 is the year organizations stop treating this as a "someday" conversation and start making actual migration decisions. Not because the technology is better elsewhere. Because the economics force the decision.
The Terraform Question: Infrastructure-as-code became the standard over the past five years. But as Terraform evolved from open-source tool to commercial product, governance, state management, and enterprise features increasingly live behind paywalls. Organizations are now asking: do we pay for Terraform Cloud/Enterprise, do we stay on open-source Terraform and build our own governance, or do we look at OpenTofu and other forks?
The "We Have AI, Now What?" Question: This is the most telling. Organizations bought AI tools because they felt they had to. They ran pilots because leadership demanded it. Now they're sitting on AI licenses, AI platforms, and AI proof-of-concepts with no clear operational framework. The technology works. But nobody owns the "how do we actually operationalize this" conversation.
These three questions share a common thread: technology decisions that seemed straightforward three years ago now require economic justification, migration planning, and operational redesign. 2026 is when those bills come due.
Art Jannicelli, Senior Solutions Architect at e360, sees infrastructure budgets getting hit by forces that have nothing to do with technology: "Tariff unpredictability and volatile RAM prices."
This is the infrastructure reality nobody talks about at conferences. You can architect the perfect hybrid cloud strategy, negotiate favorable hyperscaler contracts, and optimize your resource utilization down to the single-digit waste percentage. And then tariffs change, RAM prices spike due to supply chain disruptions, or geopolitical tensions impact hardware availability, and your carefully planned budget is suddenly off by 20%.
Organizations that built infrastructure plans assuming stable hardware economics are discovering that stability no longer exists. Server refresh cycles that used to have predictable costs now require scenario planning: what if prices go up 15%? What if lead times extend to six months? What if the specific configuration you standardized on becomes unavailable?
The enterprises that navigate this successfully will be the ones who build flexibility into procurement, maintain relationships with multiple hardware channels, and design architectures that can absorb cost volatility without requiring complete redesign.
Bryan Zanoli, Strategic Technology Advisor at e360, sees infrastructure remaining central regardless of how the "AI bubble" narrative resolves:
"If AI's growth trajectory holds, we should expect data center capacity to tighten further, hyperscaler pricing to rise, and hardware costs to remain elevated. Sustained demand will continue to drive innovation in cooling and power delivery, pushing the practical limits of modern compute density.
If the 'AI bubble' bursts, a market correction could drive down both cloud and on-prem infrastructure pricing. That type of reset may ultimately become the catalyst for a secondary, more distributed wave of AI adoption."
Think about what this means. In the boom scenario, infrastructure constraints become the limiting factor. You're competing for data center capacity, paying premium prices for compute, and dealing with power and cooling limitations that prevent you from deploying the workloads you want.
In the bust scenario, infrastructure costs fall but you still need the capacity and capability to support whatever AI adoption survives the correction. Organizations that cut infrastructure investment during a downturn often regret it when the next wave arrives.
Either way, infrastructure planning remains critical. The mistake would be assuming that the answer to "should we invest in infrastructure modernization?" depends on whether AI sustains its current trajectory. The answer is yes regardless.
Roy Douber, Senior DevOps Architect at e360, sees the specific technical debt accumulating: "Kubernetes complexity at scale: unstable clusters, brittle ingress and service-mesh setups, upgrade risk, multi-cluster sprawl, and unclear ownership across platform and application teams."
Kubernetes won the container orchestration wars. It's the de facto standard. Which means it's now mature enough to create serious operational debt when implemented poorly.
The organizations that adopted Kubernetes early and learned through painful experience have stable, well-governed clusters. The organizations that adopted it because "everyone else is doing it" without investing in platform engineering discipline now have clusters that work well enough in development and fall apart under production load.
Douber notes the specific pain points: "Insufficient testing, weak promotion controls, inconsistent environments, and too much manual 'tribal knowledge' during releases." These aren't Kubernetes problems. These are organizational maturity problems that Kubernetes exposes.
2026 is the year organizations stop treating Kubernetes as "just another technology" and start treating it as critical infrastructure that requires dedicated platform engineering teams, clear ownership models, and investment in observability and automation.
The through-line across all these predictions is clear: infrastructure economics are forcing pragmatic decisions that pure technology evaluation would never require.
Stop optimizing for the cloud pricing model of 2022. Those prices don't exist anymore. Build strategies that account for ongoing hyperscaler price increases and egress fees that make data gravity real.
Make the VMware decision. Staying is a decision. Migrating is a decision. But pretending you can defer the decision indefinitely is not a strategy.
Plan for cost volatility, not cost stability. Hardware prices will fluctuate. Supply chains will experience disruption. Build procurement strategies with flexibility, not fixed assumptions.
Invest in platform engineering for Kubernetes. If you're running Kubernetes at scale without dedicated platform teams, you're accumulating technical debt faster than you realize.
Treat infrastructure as economic leverage, not just technical capability. The best architecture is the one you can afford to run sustainably for the next three years.
The vendors will sell you on capability. The analysts will sell you on trends. But the organizations that succeed in 2026 will be the ones who make infrastructure decisions based on economic resilience, not technology fashion.
If your organization is evaluating VMware alternatives, planning data center modernization, or building economic resilience into infrastructure strategy, e360 can help.
Schedule an Infrastructure Economics Assessment