Nutanix NC2 on Google Cloud now supports Z4D bare metal with AMD EPYC

What Nutanix is adding

Nutanix says not every application takes the same path to the cloud. Some workloads can move quickly, while legacy applications, databases and performance-sensitive systems need predictable infrastructure that does not force a major redesign. Z4D support is meant to cover that second group.

NC2 runs the Nutanix software stack on Google Cloud bare metal. Teams manage workloads in their own Google Cloud account and VPC, and they use the same core data and management plane across environments, according to Nutanix. The company describes Z4D as a new infrastructure choice for workloads that benefit from direct access to physical hardware, high CPU density, local storage and low-latency execution.

Nutanix lists the following workload types as candidates:

• Business-critical databases and analytics
• Applications with demanding storage or compute needs, including virtual desktop infrastructure (VDI)
• Workloads with specific licensing or infrastructure considerations
• AI and machine learning environments that need to work alongside Google Cloud services
• Legacy applications that are difficult or costly to refactor

Z4D hardware at a glance

The Z4D bare metal configuration described for this offering combines 5th Gen AMD EPYC processors (Turin) with high-density local storage, substantial memory and high-bandwidth networking. Nutanix lists these published maximums and configurations:

• Local SSD: 84 TiB (12 × 7 TB)
• Compute: 192 physical cores (2 × 96) at 3.9 GHz
• Memory: 3,072 GiB DDR5
• Networking: up to 400 Gbps (2 × 200 Gbps)

Nutanix adds that actual availability and configuration details may vary by Google Cloud region and should be confirmed in the applicable Google Cloud and Nutanix documentation. The post does not give pricing or a regional rollout schedule.

Migration, capacity, recovery and modernization

Nutanix frames NC2 on Google Cloud as a consistent operating experience across the data center and the cloud. It names four initiatives that this consistency can simplify.

• Migrate without starting over: applications and data can move to Google Cloud with less disruption and without re-architecting every workload first. Nutanix points to legacy applications that matter to the business but were not built for cloud-native deployment.
• Extend capacity: teams can add cloud capacity for seasonal demand, geographic expansion, development and testing, or mergers and acquisitions, without a second set of infrastructure management tools.
• Improve resilience and recovery: NC2 can support cloud-based disaster recovery strategies, including Pilot-light and Zero-Compute, so protection extends from on-premises environments to Google Cloud and workloads come online when needed.
• Modernize over time: Nutanix services and Google Cloud native services, including Google Cloud Storage, BigQuery and Vertex AI, can be used together, so each application can follow its own modernization path.

What it means for hybrid cloud planning

Nutanix calls its approach workload-first. A workload might stay on premises for governance or latency reasons, move to the public cloud for capacity or geographic reach, or combine Nutanix and native cloud services as it evolves. The company says NC2 on Z4D combines Nutanix portability and operational consistency, Google Cloud's dedicated bare metal compute, the performance foundation of 5th Gen AMD EPYC, and access to Google Cloud networking and native services.

For IT teams, the practical effect is one more placement option when deciding where performance-sensitive databases, analytics or legacy systems should run, and how to connect them to Google Cloud data and AI services. Nutanix acknowledges that the right answer differs for every environment and says the value of the new option is that it gives teams more room to choose.