September 23, 2013
By Maria Minsker, contributing writer

At Oracle OpenWorld Day Two, Speed of Big Data Processing Plays Starring Role

SAN FRANCISCO—At Oracle President Mark Hurd's Monday morning keynote here, big data took center stage once more. Picking up where CEO Larry Ellison left off during his welcome address the previous day, Hurd touted Oracle's new in-memory option, and reiterated Oracle's focus on applying fast, accurate analytics to strengthen its CRM resources.

"There are nine billion devices connected to the Internet today. By 2020, that number will grow to fifty billion," Hurd said. With the explosion of those devices as well as integrated connectivity on cars and other products, big data abounds. Ninety percent of data in existence was created within the last two years, he explained, and this number will continue to grow exponentially in the coming years. "I'm not going tell you how big it's going to be. It's just going to be big," he said.

As big data grows, analytics systems have to keep up with it—not only in analysis effectiveness, but also in speed. "In the future, what will matter is not how much data you can collect, but how you can analyze it and how quickly you can do it," Fabrice Bregier, CEO of Oracle Customer Airbus, said in a video during Hurd's presentation.

Typically, it can take up to six days after a test flight to analyze data from an aircraft's 100,000 sensors, Hurd explained, but cutting that down to momentary analysis can make the cycle time six times faster. "Better analytics are faster analytics, and faster analytics save time and money," he said.

Thomas Kurian, Oracle's senior vice president of technologies development, took the stage with Hurd, and shared how the company's big data solution is speeding up data processing and improving performance. "Our Big Data Appliance is the fastest preconfigured Hadoop appliance," he said. "It can stream huge volumes rapidly into the Oracle database, where the in-memory function can scan billions of rows per second."

Kurian also introduced Oracle's Fast Data solution, a complementary solution to big data that allows users to process data in real time and take action quickly. "It's ideal for customers like Airbus, that might not be able to wait for long periods of time to process data. A pilot has to make decision instantly, so he needs to have access to data instantly as well," Kurian said.

Douglas Fisher, corporate vice president and general manager of software and services group at Intel, also touched on the speed at which data is growing, and emphasized the "need for speed" in data analytics in his presentation.

"Soon, fifty billion connected devices will generate tens of zettabytes of customer experience data. The question is, are you ready for that?" Fisher asked his audience.

Fisher discussed how Intel is working to solve the big data problem across several industries, such as healthcare. Aetna Vice President Martha Wofford took the stage with Fisher to discuss the challenges plaguing the industry—from doctor shortages to increased costs—and how big data analytics is helping to combat them. "We have to bring together data to shift the focus from worrying about healthcare to taking care of your health," she said.

Big data analytics have helped scientists take major strides in speeding up the process of genome sequencing, for example, thus paving the way for cancer researchers to find effective drugs. "Genome sequencing cost seventy million dollars in 2000. Now it's down to less than $5000. It used to take two years, and now it takes as little as two weeks. When you're dealing with people's lives, you want to eventually get it down to two hours," Fisher said. "It's not about big data. It's about big discovery."

The day's other speakers echoed the morning's sentiments, and agreed that as users continue to generate more data, businesses will need to implement the right tools to access valuable intelligence quickly for it to be useful in a fast-paced business environment.

"It's been well documented that businesses see a huge opportunity from harnessing the insights hidden within big data. It's working out how to garner those insights that is the challenge. Organizations have been hampered up to now by the need to acquire specialist skills and to build new infrastructure, by the relative immaturity of the tools, as well as by the level of business change required to sustain the value," Steve Shelton, head of data at BAE Systems Detica, said.

There is no silver bullet for businesses looking to maximize insights from big data, but as a starting point, it is of much more benefit to use a platform to deliver big data analytics as a service, Shelton said. Doing so means organizations get much greater speed to value—the insights they need quicker—with lower risk and that they are able to experiment to establish what works best for them, he explained.

"As speed to value becomes even more critical and businesses increasingly need big data analytics to be swift as well as insightful," he said, "we see big data managed platforms as being the key to unlocking value from big data over the next couple of years."