VM Sizes and Series

Choosing the right VM size is one of the most important decisions when working with Azure Virtual Machines. Azure offers hundreds of VM sizes grouped into families, each optimised for different workload types. Selecting the correct size ensures you get the performance you need without overspending.

---

Understanding VM Size Names

Azure VM sizes follow a naming convention that encodes the family, sub-family, vCPU count, and features:

Standard_D4s_v5
│        ││ │
│        ││ └─ Version (v5 = 5th generation)
│        │└── Feature flags (s = Premium Storage capable)
│        └─── vCPU count (4 vCPUs)
│
└─ Family (D = General Purpose)

Common Feature Flags

Flag	Meaning
s	Premium Storage capable
d	Local temp disk (NVMe)
a	AMD processor
p	ARM-based (Ampere) processor
l	Low memory (less RAM per vCPU)
i	Isolated (dedicated physical host)
b	Block storage performance

For example, Standard_D4ads_v5 is a general-purpose VM with 4 vCPUs, an AMD processor, a local temp disk, and Premium Storage support.

---

VM Families

General Purpose (B, D, A series)

General-purpose VMs provide a balanced ratio of CPU to memory and are suitable for most production workloads.

Series	vCPUs	RAM	Use Case
B-series	1–32	0.5–128 GiB	Burstable workloads — dev/test, small databases, low-traffic web servers
D-series v5	2–96	8–384 GiB	Enterprise applications, relational databases, mid-tier caching
A-series v2	1–8	2–64 GiB	Entry-level, development, and low-cost workloads

B-series Detail: B-series VMs accumulate CPU credits when running below baseline. These credits can be spent during bursts of activity. This makes them highly cost-effective for workloads that are idle most of the time but occasionally need full CPU power.

Compute Optimised (F series)

Series	vCPUs	RAM	Use Case
F-series v2	2–96	4–192 GiB	Batch processing, gaming servers, CPU-intensive analytics, medium-traffic web servers

F-series VMs offer a high CPU-to-memory ratio, delivering more compute power per pound. They use Intel Xeon Platinum processors.

Memory Optimised (E, M series)

Series	vCPUs	RAM	Use Case
E-series v5	2–104	16–672 GiB	In-memory databases, large caches, data warehousing
M-series	8–416	218 GiB–12 TiB	SAP HANA, extreme in-memory workloads

The M-series provides the largest memory configurations on Azure — up to 12 TiB of RAM in a single VM, purpose-built for SAP HANA and similar in-memory databases.

Storage Optimised (L series)

Series	vCPUs	RAM	Use Case
L-series v3	8–80	64–640 GiB	Distributed databases (Cassandra, Elasticsearch), data warehousing, large transactional databases