Dynamic Memory Throttling Across Heterogeneous Hyper-V Host Generations

This question addresses the runtime behavior of Dynamic Memory when a Generation 2 virtual machine undergoes Quick Migration between Hyper-V hosts with differing firmware configurations. The goal is to determine whether memory reservation thresholds, startup caps, and current allocation states are preserved, adjusted, or reset across the migration boundary without a full shutdown.

Documented Hyper-V capabilities such as Dynamic Memory real-time adjustment and Generation 2 UEFI firmware enable cost reduction for low-traffic workloads, yet the interaction between throttling mechanisms and migration triggers across heterogeneous host generations remains undocumented.

Version-dependent constraints, such as the stricter minimum and maximum RAM settings reported for Hyper-V 2012 R2, add further uncertainty to expected outcomes in mixed-generation clusters.

Specifically, it is unclear how the UEFI firmware layer of a Generation 2 VM reconciles with Generation 1 host memory management layers during a live migration event.

Under what conditions does Dynamic Memory throttling defer or reset after Quick Migration across a Gen 1-to-Gen 2 host boundary? Does the firmware version of the UEFI shell influence memory reservation continuity? Are there documented interoperability matrices for mixed-generation clusters under Live Migration with Dynamic Memory enabled?