Some industries are literally defined by data, and telecommunications is one of those industries. From apps to SIM cards to location features, data is at the core of all things mobile. But as data continues to be generated in unprecedented volumes from a variety of new sources, challenges associated with accessing that data in a timely manner are prohibitive. With traditional platforms deteriorating under the onslaught of information, a new solution, born out of gaming technology, came into being—GPU-driven analytics.
As one of the world’s largest telecommunications companies, Verizon Wireless faced the challenge of polling all of the data from the smartphones in its network of 91 million customers to assess a variety of metrics like forensics on dropped calls, sensor data, log files, customer churn, device stats, and data center performance—and the data arrives constantly and relentlessly. This query would often take hours or even days to run and more hours and days to evaluate, resulting in the company only running the analysis periodically. This led to lag times in Verizon’s ability to respond to network, customer, and company problems.
Enter GPU technology. Born out of video game and 3D technology, the GPU (or graphics processing unit) is emerging as a strong alternative to the CPU (computer processing unit). GPUs can accelerate analytics and visualization by up to 100x the speed of a CPU. Equipped with the ability to render graphics instantly, GPUs can compute, analyze, visualize, and explore massive amounts of data all on one chip.
After replacing its traditional, CPU-driven analytics platform for a GPU-driven database, Verizon is now able to run these same queries in less than a second, which has led to more frequent polling and better analytics. In addition to lightning-fast SQL queries, the company’s analysts are able to animate and visualize this data, allowing them to spot trends and anomalies instantly and intuitively.
Drawing on Verizon’s case study, Abdul Subhan and Todd Mostak dive into the challenges facing telco providers and any company that requires the processing billions of rows of mobile data. Along the way, Abdul and Todd dissect Verizon’s data analytics strategy, revealing how Verizon tackled and overcame lagging query times, campaign analysis, and mobile, real-time analytics by using the speed and transparency of GPU technology to operate more efficiently and serve customers more effectively.
Todd Mostak is the founder and CEO of MapD, a pioneer in building GPU-tuned analytics and visualization applications for the enterprise. Previously, Todd was a research fellow at the MIT’s Computer Science and Artificial Intelligence Laboratory, where he focused on GPU databases and visualization. Todd conceived of the idea of using GPUs to accelerate the extraction of insights from large datasets while conducting graduate research on the role of Twitter in the Arab Spring. Frustrated by the capabilities of conventional technologies to allow for the interactive exploration of these multimillion row datasets, Todd built one of the first GPU-based databases. Todd holds an MA in Middle Eastern studies from Harvard.
Abdul Subhan is the principal solutions architect for Verizon’s 4G Data Analytics team, where he is responsible for a developing, deploying, and managing a broad range of mission-critical reporting and analytics applications. Previously, Abdul spent close to a decade as an internal technology consultant to various parts of Verizon’s wireless, wireline, broadband, data center, and video businesses and managed data and technology with Nuance Communications and Alcatel Lucent. Abdul holds a master’s degree in computer engineering from King Fahd University of Petroleum & Minerals and an undergraduate degree in telecommunications from Visvesvaraya Technological University.
Comments on this page are now closed.
©2017, O'Reilly Media, Inc. • (800) 889-8969 or (707) 827-7019 • Monday-Friday 7:30am-5pm PT • All trademarks and registered trademarks appearing on oreilly.com are the property of their respective owners. • firstname.lastname@example.org
Apache Hadoop, Hadoop, Apache Spark, Spark, and Apache are either registered trademarks or trademarks of the Apache Software Foundation in the United States and/or other countries, and are used with permission. The Apache Software Foundation has no affiliation with and does not endorse, or review the materials provided at this event, which is managed by O'Reilly Media and/or Cloudera.