What is Azure Compute Fleet?
Azure Compute Fleet is a service, generally available since April 2025, for managing large fleets of Virtual Machines through a single deployment. Compute Fleet combines different VM sizes and purchasing models such as Spot, On-Demand, and Reserved Instances to obtain capacity quickly and optimize cost. The service can optionally distribute capacity across multiple regions.
Core Features
- Deploy up to 10,000 VMs with a single API request
- Mix of Spot, On-Demand, Reserved Instance, and Savings Plan pricing models
- Configurable allocation strategies for cost or capacity optimization
- Attribute-based VM selection (e.g., by vCPU, memory) as a preview feature
- Automatic replacement of evicted Spot VMs to restore target capacity
- Optional multi-region distribution for a single fleet (preview feature)
Typical Use Cases
Batch processing and data engineering workloads use Compute Fleet to obtain large amounts of parallel compute capacity cost-effectively. CI/CD pipelines dynamically launch build agents through a fleet. Other use cases include risk analysis, log processing, or rendering, where many concurrent instances are required.
Benefits
- Fast, consolidated access to Azure capacity across VM types
- No additional cost for the fleet service itself; only the VMs used are billed
- Automated management of Spot capacity without custom orchestration logic
- Higher resilience through distribution across SKUs, Availability Zones, and optionally regions
Integration with innFactory
As a Microsoft Solutions Partner, innFactory supports you in adopting Azure Compute Fleet: from selecting suitable VM types and allocation strategies, through cost optimization with Spot instances, to integration into existing CI/CD and data engineering pipelines.
Frequently Asked Questions
What is the difference between Compute Fleet and VM Scale Sets?
VM Scale Sets manage a homogeneous group of VMs, typically of one type. Compute Fleet combines different VM sizes, purchasing models (Spot, On-Demand, Reserved), and optionally multiple regions in a single deployment with a shared target capacity.
What happens during Spot evictions?
If a Spot VM is evicted due to price or capacity, Compute Fleet automatically tries to restore the target capacity by launching replacement VMs from the configured SKU list.
Which workloads is Compute Fleet suitable for?
Compute Fleet suits horizontally scalable, interruption-tolerant workloads such as batch processing, CI/CD build agents, risk analysis, log processing, or rendering, where many concurrent instances are needed.
What does Azure Compute Fleet cost?
There is no extra charge for Compute Fleet itself. You are billed only for the VM instances actually launched, at regular Spot or On-Demand rates, depending on VM size and region.
How many VMs can a fleet contain?
A single Compute Fleet can scale up to 10,000 VM instances, subject to additional per-region quotas. Current quota limits should be checked in the official documentation, as they can change.
Note: All product information on this page has been compiled with care, but is provided without guarantee and may be outdated or incomplete. Cloud services evolve rapidly — features, pricing, SLAs, and availability change frequently. Authoritative and up-to-date information can only be found on the official product page of Azure (official documentation). This page does not represent an offer by Azure.
