Microsoft Corporation offers a variety of Virtual Machines that data centers can choose from, giving users the opportunity for rapid development, testing, and deploying of applications in the cloud.
With multiple Virtual Machine (VM) options available on Microsoft’s Azure cloud, determining which VM fits your needs can be a daunting task.
Categorization of Azure Virtual Machines depends on sizes, types, and families. This article illustrates how Virtual Machines vary on different specifications, such as price, power, size, and functionality.
Different Azure Virtual Machines perform different mission-critical workloads. For instance, GPU-optimized VMs are ideal for GPU-intensive workloads, such as video editing and high-volume graphics rendering. CPU-intensive workloads, such as processing Big Data, are best handled by fast, high-compute CPU-optimized Virtual Machines with network interfaces.
At present, Microsoft Corporation offers six types of Azure Virtual Machines -
Determining the right VM is easy. Pick a virtual machine from these six options, the one that meets your workload.
For example, the general purpose virtual machine is an ideal choice if you are in the testing phase, hosting a small database. In case you are into memory analytics, deploying a memory-intensive database, then memory-optimized virtual machines are a perfect fit.
If you do not find a VM that meets your needs, you can browse the Azure Virtual Machine Marketplace for third-party VM images. These niche VMs, developed by third parties, fulfill workload demands not supported by stock Azure Virtual Machines.
Azure Virtual Machines are available in various sizes and families. Numbers and letters, for example, B1, NV24, D2s_v3, A1, or A0, specify the family and size of an Azure VM.
VM sizes depend on the number of central processing units (CPUs) deployed, which support different disk variations. Traditional virtual machines still use HDD disks, but most new-generation VMs deploy solid-state drives (SSDs).
Azure's Virtual Machine family comprises of varying VM sizes to suit different workload patterns. If you are uncertain of which VM size to choose, Microsoft Azure's estimators and calculators will help you decide the right size.
For instance, if you are shifting your database to Microsoft Azure, the DTU (Database Transaction Unit) Calculator will enable you to decide the size needed for your SQL database in the Azure cloud. Further, the Azure Infrastructure as a service (IaaS) cost estimation tool, developed by Microsoft, streamlines the price calculation hassle, providing you with an exact estimation of the cost of Azure migration.
With the Database Transaction Unit Calculator and IaaS cost estimation tool, you can get a correct idea of the cost of migration and the VM size needed for your workload. Feed the price and size calculator tools with figures, and the calculators will give you an estimate of the size needed by your database and the price for shifting to Azure. To compare the compute performance on Azure Virtual Machines, use the newest ACU (Azure Compute Unit) model introduced by Microsoft Corporation.
Workload-specific VM sizes, such as the NV and NC, are ideal for GPU-intensive use cases using the NVIDIA GPU card. NV and NC sizes are perfect for VDI scenarios, and for workloads that demand remote virtualization, encoding, gaming, and live streaming.
Microsoft Corporation has released for preview its latest Azure virtual machine product line, the Burstable B-Series. The new virtual machine family enables users to select a virtual machine that integrates the capacity to carry out CPU performance bursts from base level to hundred percent.
Microsoft’s Burstable B-Series virtual machines are suitable for workload models that don’t need full optimization of CPU performance throughout the day. The series is ideal for case scenarios where a server is busy within a specific time frame but remains idle for the most part of a day. The B-Series virtual machines, during the idle time, generates ‘credit’, which users can use in busy hours. Azure’s Burstable B-Series is a cost-effective deployment with performance burst options when needed.
First start low with a cost-effective virtual machine, and then scale up as and when needed. Upgrading and downgrading the size of an Azure virtual machine is possible at any point in time through Microsoft's Azure console.
Azure virtual machines integrate flexible scale sets that users can deploy to scale their applications. Besides the standard scale sets, implementation of vertical scaling is available through Azure Automation. Vertical scaling increases or decreases the size of an Azure virtual machine according to workload demands.
Costs of deploying an Azure Virtual Machine is variable. For instance, the price for an Azure A0 virtual machine, used for testing and experimenting, is 11.09 pounds per month. For an advanced virtual machine, such as Azure NV24, the cost is 3330.39 pounds per month. Deployment of an NV24 virtual machine is ideal for complex applications involving graphics rendering and Big Data analytics.
The Azure Virtual Machine pricing mentioned above are tentative price estimates for the United Kingdom. The pricing can, however, differ across regions.
There are many virtual machines within the Azure VM family with diverse pricing packages for different applications and workloads. The pricing depends on minutes/hours of usage. Users can leverage a balance in the cost of VM application by going online or offline as needed.
Microsoft Virtual Academy (MVA) incorporates the latest curriculum on Azure Virtual Machines, related to VM families, products, pricing, and features. The curriculum gives users a hands-on experience of Azure VM functionalities, before their integration into Microsoft Official Curriculum (MOC).
Enterprises willing to migrate to Azure Virtual Machine can consider pursuing Microsoft’s certification on Azure VM infrastructure and solutions, which encompasses the creation, implementation, and management of Azure Virtual Machines. Azure VM certifications will help build knowledge and experience on migration and deployment of Azure VMs.