Question 1

How much can a typical AVD environment realistically save?

Accepted Answer

In our cost model, an estate with standard business hours and no prior optimization recovers 30–60% of compute spend from demand-based scaling and scheduled power management (the calculator defaults to the conservative end). Storage tiering and right-sizing are modeled separately on top. The full methodology is published on the AVD cost calculator.

Question 2

Does auto-scaling disconnect active user sessions?

Accepted Answer

No. Atlas-automated scale-in puts a host into drain mode, re-reads the live session count from Azure immediately before deallocating, and abandons the action if any session (active or disconnected) is present or the count cannot be read. It never forces a logoff and has no timeout that escalates to one. Manual power actions and Azure outages are outside that guard.

Question 3

What VM sizes work best for knowledge worker AVD workloads?

Accepted Answer

D4as_v5 (4 vCPU, 16 GB) handles 4-6 concurrent knowledge worker sessions comfortably in multi-session host pools. D2as_v5 is appropriate for lighter task-worker profiles. D8as_v5 should be reserved for power users or workloads with higher CPU or memory demands.

Question 4

How quickly does Atlas AVD deliver cost savings?

Accepted Answer

Scaling policies begin deallocating empty hosts as soon as they are enabled, so the effect shows in the next Azure billing period. In our model the largest share comes from overnight and weekend idle hours.

Question 5

Can Atlas AVD optimize costs across multiple Azure subscriptions?

Accepted Answer

Yes. Atlas AVD's multi-subscription management supports cost tracking, scaling policies, and recommendations across all AVD environments in a tenant, with per-subscription or aggregate reporting.

Azure Virtual Desktop Cost Optimization