Microsoft is making it easier for enterprises and database administrators (DBAs) to harness the cloud to power their cutting-edge, data-intensive workloads.
SQL Server 2017 already ships with built-in AI (artificial intelligence) capabilities, as well as Linux and DevOps-friendly application container support. Microsoft is expecting many of its customers to run the database software and put those capabilities to work in their own data centers.
But the software giant is also making room on its Azure cloud computing platform for those businesses that may one day, perhaps inevitably, want to tap the power of Microsoft's hyperscale cloud to modernize their operations in the face of massive data growth. One of the key areas driving digital transformation at the software giant's customers is "data plus AI," John "JG" Chirapurath, general manager at Microsoft's Data Platform unit told eWEEK.
"What we've noticed over a period of time is that it's inescapable—in the world of Hadoop and [other big data platforms]—that we're collecting more data than ever before," he continued. "When you add the innovations that you've seen in AI—Azure ML [Machine Learning] is an example of that—coupled with the immense computing power that's available in the cloud, what we're seeing is customers transform their [businesses] across the data and AI spectrum in remarkable ways."
At Jet.com, those ways include personalized pricing offers delivered in real-time, even at times of peak demand. Meamwhile Rockwell Automation is using Azure to correlate weather conditions and other metrics to enable predictive maintenance on its IoT-enabled climate-control equipment.
All types of businesses can take advantage of Microsoft's Data Platform, a hybrid-cloud based approach to storing, managing and deriving more value from data, said Chirapurath.
During PASS Summit 2017, which runs Oct. 31-Nov. 3 here in Seattle, Microsoft executives showed off some of the new and forthcoming cloud-based capabilities that can help organizations make the most out of the enormous volumes of data they generate. In demonstrations, Azure SQL Database soaked up 1.4 million rows per second. Using the T-SQL (Transact-SQL) programming extension to run Azure ML analytics models, customers can load large amount of data in little time, or at an average of 20 milliseconds per row of data.
Microsoft is also working on the release of hybrid data integration capabilities for Azure SQL Data Warehouse. Currently in preview, they will enable users to run SSIS (SQL Server Integration Services) workloads without making any changes to SSIS packages, enabling data integration that spans both on-premises systems and the cloud.
For faster overall performance, Microsoft is currently running a public beta of its new Compute Optimized Tier for Azure SQL Data Warehouse. Offering 100 times the query performance and five times the scalability of its predecessor, it can help businesses speed up big data analysis. It also boasts query speeds of up to 4.5 billion rows per second and scan throughput of up to 360 GB per second.
The new performance tier enables users to provision more than 4,000 virtual CPUs in minutes. The new Compute Optimized Tier is also the first cloud-based data warehouse that uses NVMe SSDs for speedy storage performance, according to Microsoft.