Virtual machine performance degradation after vMotion to Dell PowerEdge servers with Intel Xeon Granite Rapids

Products

• VMware vSphere ESXi

Issue/Introduction

Symptoms

• Virtual machines (VMs) experience performance degradation or high CPU utilization on the guest OS after being migrated via vMotion to Dell PowerEdge servers equipped with Intel(R) Xeon(R) Granite Rapids processors.
• This issue is more prominent when migrating workloads from older CPU architectures (e.g., Ice Lake).

Environment

VMware vSphere ESXi 8.0 Update 3e and later

Dell PowerEdge servers with Intel Xeon Granite Rapids processors (e.g., R670, R760)

Cause

According to Dell technical documentation, 17th Generation PowerEdge servers with Intel Granite Rapids processors require specific BIOS-level optimizations to achieve maximum performance.
If these recommended settings (LOM/SNC) are not applied, the system may not fully leverage the new CPU architecture, leading to increased processing latency for virtualized workloads

Resolution

As this behavior is related to hardware-level optimization, please contact Dell Technical Support to confirm if the following BIOS settings should be enabled for your specific server configuration and workload:

Consult with Dell Technical Support Contact your hardware vendor to verify the applicability of the following features:

• Latency Optimization Mode (LOM)
• Sub-NUMA Clustering (SNC)

Refer to official Dell documentation for further details: Dell customized VMware ESXi support on the 17th generation of PowerEdge server platforms

Additional Information

• If using vSAN, please also refer to KB 393469 for specific guidance on Intel Xeon 6700/6500-Series processors.
• Intel Xeon Granite Rapids を搭載した Dell PowerEdge サーバーへの vMotion 後の仮想マシンのパフォーマンス低下