Updated Analytics and Big Data Comparison: AWS Vs. Azure

Building upon my earlier post, today, I wanted to share with you updated graphics and links for the analytic and big data services offered by Microsoft Azure and Amazon Web Services.

It is my hope that this post will be a starting guide for you when you need to research these analytics and big data services. I have included relevant links for each service, along with some commentary, in the text of this post below. I’ve done my best to align the services, but there is some overlap between offerings. 

I’m not going to do a feature comparison here because these systems evolve so quickly that I’d spend all day updating the info. Instead, you get links to the documentation for everything and you can do your own comparisons as needed. I will make an effort to update the page as frequently as possible. Let's get started.

Data Warehouse

Azure offerings: SQL Data Warehouse

AWS offerings: Redshift

It feels like these two services have been around forever. That’s because, in Internet years, they have. Redshift goes back to 2012, and SQL DW goes back to 2009. That’s a lot of time for both Azure and AWS to learn about data warehousing as a service. 

Data Processing

Azure offerings: HDInsight

AWS offerings: Elastic MapReduce

Both services are built upon Hadoop, and both are built to hook into other platforms such as Spark, Storm, and Kafka.

Data Orchestration

Azure offerings: Data Factory, Data Catalog

AWS offerings: Data Pipeline, AWS Glue

These are true enterprise-class ETL services, complete with the ability to build a data catalog. Once you try these services, you will never BCP data again. 

Data Analytics

Azure offerings: Stream Analytics, Data Lake, Databricks

AWS offerings: Lake Formation, Kinesis Analytics, Elastic MapReduce

I didn’t list Event Hubs here for Azure, but if you want to stream data, you are likely going to need that service as well. And Kinesis is broken down into specific streams, too. (In other words, “Analytics” is an umbrella term, and is one of the most difficult things to compare between Azure and AWS). 

Data Visualization

Azure offerings: PowerBI

AWS offerings: QuickSight

I saw some demos of QuickSight while at AWS re:Invent last fall, and it looks promising. It also looks to be slightly behind PowerBI at this point. Of course, we all know most people are still using Tableau, but that is a post for a different day. 

Search

Azure offerings: Elasticsearch, Azure Search

AWS offerings: Elastisearch, CloudSearch

Elastisearch for both is just a hook to the Elastisearch open-source platform. For Azure, you have to get that from their marketplace (that’s what I link to because I can’t find it anywhere else). One of the biggest differences I know between the services is the number of languages supported. AWS CloudSearch claims to support 34, and Azure Search claims to support 56.

Machine Learning

Azure offerings: Machine Learning Studio, Machine Learning Service

AWS offerings: SageMaker, DeepLens

DeepLens is a piece of hardware, but I wanted to call it out because you will hear it mentioned. When you use DeepLens, you use a handful of AWS services such as SageMaker, Lambda, and S3 storage. I enjoyed using Azure Machine Learning Studio during my data science and big data certifications. But the same thing is true, you use associated services. This makes price comparisons difficult.

Data Discovery

Azure offerings: Data Catalog, Data Lake Analytics

AWS offerings: Athena

Imagine a library without a card catalog and you need to find one book. That’s what your data looks like right now. I know you won’t believe this, but not all data is tracked or classified in any meaningful way. That’s why services like Athena and Data Catalog exist.

Pricing

Azure Pricing calculator: https://azure.microsoft.com/en-us/pricing/calculator/

AWS Pricing Calculator: https://calculator.aws/

Same as the previous post, you will find it difficult to conduct an apples-to-apples comparison between services. Your best bet is to start at the pricing pages for each and work your way from there.

Summary

I hope you find this page (and this one) useful for referencing the many analytic and big data service offerings from both Microsoft Azure and Amazon Web Services. I will do my best to update this page as necessary and offer more details and use cases as I am able.