Modern IT estates have traded single-stack simplicity for a mix of clouds, hardware, and deployment models. The Register's sponsored feature from Nutanix says that mix now contains a growing split between VMs and containers: AI workloads often go straight to containers, while the data and enterprise applications AI needs remain on VMs. The result is a "VM-container divide" that complicates operations, especially for agentic AI.
Running separate VM and container stacks brings practical problems, according to the article: duplicate capacity plans, separate support contracts, fragmented governance, and inefficient resource use. VM clusters sit underutilized while container pools are overprovisioned, leaving installed capacity stranded unless moved manually. Nutanix survey data cited in the piece says 85% of IT executives see AI meaningfully accelerating container adoption, with 29% seeing a major acceleration.
Nutanix's answer is a "dual-native" architecture that treats VMs and containers as equal citizens in one operating environment, so agents can call existing VM-based applications and data without migration. The source is a single sponsored feature, so the claims represent Nutanix's position rather than an independent benchmark.