Can I Use Reserved Instances with AVD Scaling?
A common question we get: “If I use Reserved Instances, do I still need scaling?”
The answer is yes — and here’s why they work brilliantly together.
What Are Reserved Instances?
Azure Reserved Instances (RIs) let you commit to 1 or 3 years of VM usage in exchange for significant discounts (typically 30-60% off pay-as-you-go).
The catch? You pay whether you use them or not.
The Smart Approach: Baseline + Burst
The ideal AVD setup combines both:
Baseline (Reserved Instances)
Your minimum daily usage. For many organisations, this might be 20-30% of peak capacity.
Example: If you need 50 VMs at peak, perhaps 15 VMs are always in use during business hours.
Reserve these 15. Get the discount.
Burst (Pay-as-you-go + Scaling)
Everything above baseline scales up and down with demand.
Example: The other 35 VMs spin up when needed, shut down when not.
The Maths
| Approach | Annual Cost |
|---|---|
| 50 VMs, 24/7, pay-as-you-go | £36,000 |
| 50 VMs, 24/7, all reserved | £18,000 |
| 15 reserved + 35 scaled | £12,000 |
Best of both worlds wins.
How The Smart Scaler Fits In
Our scaling works perfectly with RIs:
- We track which hosts are Reserved vs pay-as-you-go
- Reserved hosts are preferred for baseline load
- Pay-as-you-go hosts scale up/down
- You maximise both discounts
Tips for Getting It Right
- Analyse before reserving — understand your actual baseline first
- Start conservative — reserve less than you think (you can add more)
- Right-size first — no point reserving oversized VMs
- Review annually — usage patterns change
Need help understanding your baseline? Our historical analytics show exactly how many hosts you consistently need.