Tuesday, March 19, 2024

Microsoft: Azure Batch Review

Datamation content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.

Microsoft’s Azure Batch provides burst, scheduled, and on-demand access to cloud computing resources hosted on the Azure cloud platform. 

As data scientists and application managers encounter resource constraints, they can turn to Azure Batch to deploy computers and clusters rapidly through automated processes.

See below to learn all about Azure Batch and where it stands in the compute market:

See more: The Best Data Center Companies

Microsoft Azure Batch and the compute market

Microsoft’s Azure Batch competes in the cloud computing market. Allied Market Research estimated cloud computing revenues of $325 billion in 2019, with projections to reach $1.6 trillion in 2030, at a compound annual growth rate (CAGR) of 15.8%.

Grand View Research estimated cloud computing revenues of $369 billion in 2021, with projections to reach $1.5 trillion, at a CAGR of 15.7%. Meanwhile, Markets and Markets estimated a $445.3 billion market size in 2021, with projections to reach $947 billion by 2026, at a CAGR of 16.3%.

For Q4 2022, Microsoft reported $20.9 billion in intelligent cloud revenue and 22% increases in server products and cloud services. Within the marketplace, Microsoft is estimated to hold a 21% market share and is ranked No. 2 within the cloud services market.

The major competitors in the cloud computing market include:

  • Adobe 
  • Alibaba Group
  • Amazon Web Services (AWS)
  • Cisco 
  • Dell Technologies
  • Hewlett Packard Enterprise (HPE)
  • IBM 
  • Google Cloud
  • Oracle
  • Rackspace Hosting
  • Salesforce 
  • SAP
  • Workday

Microsoft Azure Batch key features

  • Job scheduling engine as-a-service
      • Quickly deploy or scale virtual machines (VMs)
      • Deployable infrastructure in up to thousands of machines
      • Identify failures and re-queue work
  • Plan and scale workloads
      • Stage data
      • Execute compute pipelines
      • Autoscale on work in queue
      • Cloud-enable batch and high-performance computing (HPC) applications
      • Does not require engineers to queue, dispatch, or monitor workloads
      • Batch process jobs on demand
      • Integrate with applications
      • Scale the pool of virtual machines as jobs launch or finish
      • Control scale to meet deadlines and manage costs
      • Automated Lab policies to shut down or start up virtual machines
  • Embedded security
      • Supports Windows Virtual Desktop Connections
      • Incorporates GitHub Codespaces
      • Links to Azure DevTest Labs
  • Flexible infrastructure 
      • Linux: CentOS, Ubuntu, SUSE Linux Enterprise
      • Windows OS
      • Microsoft .NET support
      • Docker containers
      • Support for development tools such as Python and Java
      • Batch Blender animation jobs
  • Pay only for consumed resources

See more: How Data Centers are Used by Bosch, PayPal, Groupon, Orange, and Suez: Case Studies

Microsoft Azure Batch key benefits

When selecting a cloud computing solution such as Azure Batch, organizations seek to deploy massive computing power in a short period of time. This provides a cost effective way to obtain key benefits related to:

Animation rendering

While modern animation involves heavy computing skills, most studios want their artists focused on developing the artwork and not getting lost in the complexities of managing computing resources. In selecting a solution such as Azure Batch, animators can reduce the time needed to complete jobs and provide the scale to securely collaborate with global teams.

Batch incorporates professional graphics programs as well as Blender, a free open-source 3D animation suite with tools for game creation, modeling, rigging, and simulation and motion tracking. Burst computing allows for faster rendering, whether done fully in the cloud or using Azure FXT Edge Filer to combine local and cloud resources.

Burst computation needs

To manage the natural ebbs and flows of business computing needs, elastic computing solutions can deploy instances to address sudden and temporary increases in computing. With a variety of supported operating systems and computing options, Microsoft Azure Batch can help provide the right computing power at the right time.

Unlimited scale

In addition to quick deployments for burst needs, Azure Batch can deploy at scale. Using clusters and instance options, an organization can easily deploy hundreds of thousands of cores to tackle the largest computing projects with ease.

Microsoft Azure Batch use cases

Cloud9 

The esport company Cloud9 wanted to deploy a video review tool (VRT) for Valorant to train their players for improved performance. The computer vision models and gameplay video generated long-term jobs that needed to be managed in Docker containers and consumed huge computer resources.

Cloud9 selected Azure Batch for its ability to deploy on demand and as needed.

“We needed Batch, because VRT jobs are very long running,” says Halee Mason, lead data scientist, Cloud9. 

“We have a video analyzer pool that is always pre-warmed and waiting for new jobs to queue, and the Azure Batch job is where all the magic happens.”

Geninus

The Samsung Genome Institute spun off its medical genomic analysis company, Geninus, to commercialize sequence-based cancer panels. Geninus needed burst computing power supported by compatible computing resources to perform genome analysis on its patients and develop artificial intelligence (AI) algorithms.

“We utilized Microsoft Azure Batch technology to meet the intensive computing tasks required for genome sequencing without worrying about infrastructure management while minimizing operational costs,” sayd Donghyun Park, director, Geninus. 

“With Microsoft Azure SQL Database and Microsoft Azure Storage, we’ve been able to handle massive amounts of data quickly and cost-effectively on a scale that’s hard to operate with internal on-premises infrastructure.”

Kensington Mortgages

Kensington goes beyond income and employment records to determine lending criteria for new customers. Kensington needed more computing power to support its efforts to analyze more complex income circumstances for self-employed, small sole proprietors, and other customers difficult to analyze for the general financial market.

Kensington selected Amazon Batch as a key tool to coordinate the computers to run their proprietary algorithms on their large datasets.

“Everything is in one place,” says Sughasini Ravikumar, senior analytics solutions architect, Kensington Mortgages. 

“Performance optimization is so easy with Azure Batch and supporting Azure platform services.”

Microsoft Azure Batch differentiators

Application development

While most cloud environments support application development, many of them are forced to figure out integrations with Microsoft-owned tools, such as Microsoft Windows or GitHub. In addition to inherent integration with key components, Microsoft also offers a host of tools available for application development built into the ecosystem, such as Azure Load Testing, Azure Chaos Studio, Azure DevTest Labs, and more.

Automated resource scaling

Azure Batch ties into cluster job schedulers or integrates with software-as-a-service (SaaS) tools to automatically provision computing power. The tight integration backed by Microsoft security and Azure infrastructure provides automated as-needed and secure computing power with minimal or no human interaction.

Amazon Batch will ensure that the deployed resources meet requirements, such as encryption, and can identify failures and re-queue work as needed. Customers can monitor and refine the requirements, and Amazon Batch will adjust deployments to meet operational, budgetary, and security goals.

Microsoft ecosystem

While plenty powerful simply as a brand, Microsoft goes further and allows organizations to tap into a fully developed Microsoft ecosystem of compatible applications and certified partners. Azure Batch easily integrates with other Microsoft products, such as Active Directory, Azure DevTest Labs, GitHubCodespace, Windows Virtual Desktop, and more.

In the Azure store, customers can select from a number of third-party vendors that develop Azure virtual computing compatible tools to run on or to connect with Azure Batch. For organizations that need additional support, Microsoft has a number of partners that can provide consulting or full outsourcing services related to Azure Batch.

Queue management

Customers can use Azure Batch to develop workload, data, and application queues that determine where data flows, how workloads and data should be distributed, and the commands and parameters for executing each stage of the queue. Batch allows data to be shared between steps and for centralized management of the entire execution.

Batch will identify job failures and automatically re-queue work according to the established scale and workload parameters established in the software. Data managers can control costs, meet deadlines, and scale automatically without huge demands on their time.

User reviews of Microsoft Azure Batch 

While several user review sites offer reviews of Microsoft Azure, a couple sites show reviews for the more specialized Azure Batch technology.

User Review Site Rating
TrustRadius 8.5 out of 10
G2 4.0 out of 5

Microsoft Azure Batch pricing

As with most cloud computing applications the price depends heavily upon the options selected. Customers will need to navigate options, such as:

  • OS selection: Linux VM, Windows VM, or Cloud Services
  • Global region
  • Government or commercial hosting
  • With or Without Azure Hybrid Benefit pricing
  • Type of need, including general purpose, memory intensive, network optimized, compute intensive, latest generation hyper-threaded chips, and high memory to core ratio
  • Computing options for cores, RAM memory, and disk space
  • Pricing rates (hourly, monthly)
  • Payment options, such as pay-as-you-go spot priority, pay-as-you-go low priority, pay-as-you-go normal priority, one year reserved, and three year reserved

Not all options apply to all instances. Additional fees may apply for graphics and rendering application licensing available hourly or monthly for software, such as Autodesk Maya or Chaos Group V-Ray RT.

Microsoft commonly offers bulk discounts, and pricing discounts from these listed prices are often available through Microsoft resale partners.

Conclusions

When pushing processes, workloads, and applications to the cloud, organizations can use the elastic capabilities of cloud computing to manage costs and still deliver on intense computing needs. Microsoft Azure Batch helps data and application managers schedule, share, and route workloads in an automated fashion without heavy server management skills. When developing a cloud strategy, Microsoft Azure Batch should be considered and tested as a solution for anyone needing flexible and automatable compute power.

See more: The Edge Data Center Market

Subscribe to Data Insider

Learn the latest news and best practices about data science, big data analytics, artificial intelligence, data security, and more.

Similar articles

Get the Free Newsletter!

Subscribe to Data Insider for top news, trends & analysis

Latest Articles