More Choices for Hybrid Cloud: New Intel Xeon 6 Instances for Nutanix Cloud Clusters on AWS

Migration Doesn't Come in One Size

Every migration plan runs into the same fact sooner or later: some workloads don't line up neatly. A database wants memory. An analytics job wants cores. A file tier wants capacity. Some applications carry compliance requirements that decide where and how they run, while others could move tomorrow if there were somewhere ready to put them.

Customers also want to move at their own pace. Very few IT teams can stop, re-platform everything, and start over on a new set of tools. They want to shift workloads in phases, keep one operating model across the datacenter and the cloud, and make each move when the business is ready for it.

Infrastructure choice is what makes that kind of plan workable. When the hardware options match the workload mix, the migration plan can follow the applications instead of forcing them into whatever capacity happens to be available.

At AWS re:Invent, we'll be showcasing how NC2 on AWS provides customers with hardware options, including new Intel Xeon 6 instances, to match the workload mix. This allows migration plans to follow the applications instead of forcing them into whatever capacity happens to be available.

Now Available: Intel Xeon 6 Instances for NC2 on AWS

Nutanix Cloud Clusters (NC2) on AWS now supports six additional Amazon EC2 bare metal instance types built on Intel Xeon 6 processors: c8id, m8id and r8id, each in a metal-48xl and a metal-96xl size. That gives NC2 customers a compute-optimized, a general-purpose and a memory-optimized option at two sizes, all running the same AOS and AHV stack they run on-premises.

What the New Instances Give Customers

The first benefit is fit. Teams can align each NC2 deployment to the requirements of the workloads it hosts and to where those workloads sit in the migration plan. A memory-bound tier and a compute-bound tier no longer have to share one compromise configuration.

The second is pacing. A phased migration to AWS needs landing capacity that matches each wave of applications, and more instance choices make it easier to size each wave on its own terms.

The third is resilience in planning. When on-premises hardware procurement is delayed, NC2 on AWS gives teams a place to run the same stack while they wait, and it helps reduce their exposure to hardware price and supply-chain volatility.

Through all of it, the operating model stays the same. Clusters on-premises and on AWS run AHV and are managed through Nutanix Prism Central, so the policies, workflows, and skills a team already has, carry over. The supported instance types are validated configurations, which can help remove a good deal of uncertainty from capacity planning.

Intel and AWS

For customers, the value of the partnership shows up as choice without a change in how they work. They pick the AWS bare metal instance that suits the workload and run it with the same Nutanix operating model they use everywhere else.

“Customers shouldn’t have to choose between infrastructure flexibility and operational consistency as they move workloads to the cloud," said Srini Krishna, Fellow at Intel Data Center Group. "Intel Xeon 6 gives organizations a powerful, flexible foundation to right-size infrastructure for diverse workload needs, while our collaboration with Nutanix and AWS helps extend a consistent operating model from the data center to the cloud.”

The c8id, m8id and r8id instances bring the Intel Xeon 6 generation to NC2 on AWS, in compute-optimized, general-purpose and memory-optimized shapes.

"At AWS, we're focused on giving customers the right instance for every workload. Intel Xeon 6-based Amazon EC2 C8id, M8id, and R8id instances are now available on NC2 on AWS, giving Nutanix customers the ability to match compute, memory, and storage to each workload. With up to 15% better price-performance, 2.5x more memory bandwidth than the previous generation, and significant gains across workloads like PostgreSQL, NGINX, and AI deep learning, customers can rightsize their infrastructure and lower their cloud costs."
— Art Baudo, Head of Product Marketing, Amazon EC2, AWS

Technical Details

Supported Instance Types

NC2 on AWS supports the following bare metal instance types, grouped by purpose.

The Xeon 6 instances require AOS 7.5.1.x or later with AHV 11.0.1. Types marked * require AOS 7.5.1.x or later. i3.metal is a legacy type: existing clusters can keep it, but it can't be used for new clusters, hibernation or cluster expansion.

The six Xeon 6 instances in detail:

Availability varies by AWS region. The NC2 on AWS deployment and user guide lists the current region matrix for each instance type.

Where EBS Fits

Figure 1 – EBS support across Intel EC2 instances

The Xeon 6 instance types are local-NVMe only. Cluster data lives on the NVMe drives inside each node, and the way to add capacity is to add nodes.

Every NC2 node, Xeon 6 included, boots from two encrypted gp3 EBS volumes: 100 GB for AHV and 150 GB for the Controller VM. Those are boot volumes, not storage capacity for workloads.

Optional EBS capacity, the Cloud-SSD tier, is a separate thing. It attaches to i3en, i4i (AOS 6.8 and later), i7ie and i7i (AOS 6.10.1.5 and later), and m5d and g4dn (AOS 7.3 and later). It's available for new, homogeneous clusters with NCI Ultimate or AOS Ultimate licensing. It can't be attached to c8id, m8id, r8id or to any of the other types that arrived with AOS 7.5.1.x. The diagram above shows where each tier sits.

Heterogeneous Clusters

A single NC2 cluster can mix up to three instance types, either at creation or when expanding a running cluster, subject to AWS bare metal availability in the region. The supported combinations are:

1. i3en, i4i, i7ie-24xl, i7ie-48xl, i7i-24xl and i7i-48xl
2. z1d, m5d and m6id
3. m8id, r8id and c8id at the 48xl size
4. m8id, r8id and c8id at the 96xl size
5. m6id with m6idn

Mixing within the Xeon 6 family is size-specific. Any of the three 48xl types can share a cluster, and any of the three 96xl types can share a cluster, but 48xl and 96xl nodes don't mix. A cluster with different workload profiles can pair memory-heavy r8id nodes with compute-heavy c8id nodes in the same size group and place each workload where it fits best. Expansion happens in multiples of the cluster's replication factor, and g4dn.metal stays homogeneous-only.

Moving Workloads Across CPU Generations with Advanced Processor Compatibility

A hybrid estate almost always spans CPU generations. The on-premises cluster may be a few years old, while the new NC2 cluster runs Xeon 6. AHV handles this in two ways.

By default, AHV gives each VM the highest CPU generation common to every node in the cluster. Adding a node from a different generation can change that baseline, and some VMs may need a power cycle to pick it up. The system notifies you when that's the case.

Advanced Processor Compatibility gives administrators direct control. You pin a VM to a named CPU generation the cluster supports. The Intel options span Intel Xeon E5 processors (v1 through v4) and 1st through 5th Gen Intel Xeon Scalable processors (4th and 5th Gen share a single option), with Intel Xeon 6 processors added in AHV 11.0.1. Enabling the feature without naming a generation defaults to Intel Xeon E5 v4. The pinned generation stays with the VM through power cycles, migrations and HA restarts, and a cluster only offers generations that all of its nodes support.

The choice of generation is a tradeoff. Pinning to the oldest generation in the estate gives the widest live migration reach across clusters. Pinning to the newest gives the VM the most CPU features.

Live migration works when the destination supports the CPU features the VM was given. With VMs pinned to a generation both sites support, a VM can move from an older on-premises cluster to a Xeon 6 NC2 cluster on AWS and back again. Advanced Processor Compatibility requires AOS 6.8 or later, AHV 20230302.100173 or later, and Prism Central 2024.1 or later.

What’s Next

Together, AWS, Intel, and Nutanix give customers the infrastructure choice and operational consistency to match each workload, migrate at their own pace, and run the same Nutanix operating model across the datacenters and AWS.

Nutanix continues to evaluate support for additional instance types based on customer requirements and technical validation. To tell us which models matter most for your migration, or to ask about early access, contact your Nutanix sales team. To check regional availability and plan a deployment today, start with the NC2 on AWS deployment and user guide.

Nutanix is a Gold Sponsor of AWS re:Invent 2026, November 30 to December 3 in Las Vegas. Stop by booth 1868 to talk through NC2 on AWS and the new Xeon 6 instances with the team.

Resources

• Learn more about hybrid cloud deployments with NC2 on AWS
• Take a Free Test Drive of NC2 on AWS
• Explore Cost Savings and Benefits of NC2 on AWS
• Peco Foods: Strengthening Business Continuity and Accelerating DR
• Read the Intel Hybrid Cloud Solution Brief
• Migration Promotion for VMware, Red Hat, & Oracle Customers

©2026 Nutanix, Inc. All rights reserved. Nutanix, the Nutanix logo and all Nutanix product and service names mentioned are registered trademarks or trademarks of Nutanix, Inc. in the United States and other countries. All other brand names mentioned are for identification purposes only and may be the trademarks of their respective holder(s).

©Intel, the Intel logo and other Intel marks are trademarks of Intel Corporation or its subsidiaries.