PORTLAND, Ore.—Unless you've been living off the grid, you already know that the world is drowning in a tsunami of Big Data, especially in manufacturing settings where realtime analytics can pinpoint semiconductor fabrication errors and processing inefficiencies that could spell the difference between success and failure.
However, by integrating sophisticated analytics into Microsoft's structured query language (SQL), Advanced Micro Devices Inc. (AMD) and Samsung Electronics Co. Ltd. report a 90 percent reduction in data warehousing support efforts and a 10-fold boost in the speed of energy-efficiency analytics, respectively.
"Microsoft is leveraging its business intelligence (BI) expertise to offer Big Data solutions for manufacturing, including the automotive and electronics industries," said Sanjay Ravi, worldwide managing director for discrete manufacturing at Microsoft. "We believe that the confluence of ubiquitous sensors, connectivity, on-demand cloud processing and realtime analytics uniquely positions Microsoft for a new era of instant actionable BI."
One of Microsoft's biggest design wins so far is in the automotive industry, where it convinced Ford to adopt its Windows Embedded Automotive solution for Ford's Sync platform. Sensor data from the automobile itself, along with infotainment- and user-generated queries, can be integrated by SQL both in-vehicle and using Microsoft's cloud computing environment, Azure, to provide seamless automotive analytics and realtime intelligence for everything from malfunctions to voice recognition.
Microsoft's Windows Azure cloud computing platform and its SQL Server with StreamInsight aim to streamline manufacturing analytics.
"The underlying enabler for Ford Sync's connected vehicle capability is the Microsoft's embedded automotive platform that actually sits in the vehicle capturing information from sensors and the driver to deliver an intuitive consumer experience," said Ravi.
In the electronics manufacturing market, Microsoft's Big Data capabilities are changing the rules of the game in what it calls the "democratization" of process control, energy efficiency and other realtime manufacturing information. In the past, raw manufacturing data from sensors was merely streamed to a passive data warehouse, which was only consulted by employees when problems arose—to determine what caused the problem—or to produce off-line reports about the efficiency of past manufacturing runs. For the future, however, Microsoft is aiming to streamline manufacturing with analytics that evaluate data as it is collected to head-off problems before they become costly, and to enable employees anywhere in a company to visualize operations in realtime, using its Hadoop-based distribution on Windows Azure and SQL's StreamInsight analytics.
At AMD, for instance, Microsoft's SQL has been able to institute a 90 percent cut in the previous efforts by employees tracking semiconductor wafer testing data. By running Microsoft's Parallel Data Warehouse software on AMD's massively parallel processing (MPP) Opteron processors, employees are now able to track terabytes of wafer test data in near realtime, thereby heading-off problems before they result in whole batches of faulty wafers.
And at Samsung, an even bigger data deluge—up to 80 terabytes—is now being processed by HP ProLiant servers running Microsoft's SQL and Fast Track Data Warehouse software to analyze energy efficiency at Samsung's fab in Hwasung (South Korea). As a result, Samsung reports that its analytics now run 10-times faster on datasets that are 10-times larger, despite the fact that it now only needs 75 percent as much computing power than before.