Hi Greco Alessia,
The technically correct and supported approach in a QEMU ARM benchmarking environment is to stabilize the virtualization layer. This includes keeping a fixed virtual CPU topology, pinning QEMU vCPU and IOThread execution to dedicated host cores, minimizing host scheduler interference, and configuring the Windows guest for maximum performance using documented powercfg processor settings such as PROCTHROTTLEMIN 100 and PROCTHROTTLEMAX 100. These policies are associated with the active power scheme stored under HKLM\SYSTEM\CurrentControlSet\Control\Power\User\PowerSchemes, but they influence processor policy behavior rather than overriding runtime telemetry.
There is currently no supported Microsoft mechanism that forces Windows 11 ARM to permanently trust static SMBIOS or ACPI CPPC frequency declarations over live processor performance feedback in a virtualized environment. Remaining clock variability after host-side stabilization is typically caused by hypervisor scheduling behavior, thermal management, or runtime power-management heuristics rather than missing firmware fields.
Hope this answer brought you some useful information. If it did, please hit “accept answer”. Should you have any questions, feel free to leave a comment.
VPHAN