“Federal Tech Talk” was joined by guest Tim White, Global Head for Government & Intelligence at YarcData, a Cray Company, “a super-computing company,” on Federal News Radio 1500 AM last week.
Host John Gilroy chatted with White on issues such as IT trends, challenges agencies might face with the public and private sectors, the concept of big data, the power of visualization, the technologies which help complement big data, and more.
With a mission statement of, “well known brand with a brand new mission,” White said in the interview that, “it’s an exciting time to be at YarcData. We feel like we’ll be able to provide a new way to look at data.”
White credits his experience in the military (for 20 years, of which 13 were in active duty), education, and medicine fields, to having a multidisciplinary approach to big data.
YarcData uses XMT technology and the appliance URiKA. XMT, the chip that goes into the URiKA appliance. URiKA is a graph analytics appliance and a “hypothesis generating engine.” It can be used to manipulate large amounts of data, and White describes YarcData as the slotmaschinen spielen engine behind URiKA to deliver the visual analytics, most commonly in graph form.
White, who believes that everyone is now becoming an IT expert because you need to access it and use it,” also discusses the bottleneck, as it is called, on big data, most specifically regarding the input and output of data, and just what big data is to him. In an abstract sense, White said big data is “looking at commodity hardware and pushing the leading edge.” But he said that most users don’t have much data so the problem is not that large—it’s in fact, relative.
“Problems in big data are the same, just at a different scale,” White said. “It’s something that overwhelms the structure and the system and its processing capability. It’s a relative term.”
Gilroy and White also discussed problems facing big data, touching upon issues and concerns of replication of data, and what White called “making the move from a centralized architecture into one that’s distributed across a cloud.”
Visualization is labeled as key for big data, with White saying that one must “shape analytics against hardware, not against the problem” so that “the hardware matches the problem.”
The “big four” in big data are discussed (volume, variety, velocity, voracity), as well as the impending importance of social technology–one billion smartphones and 400 million tablets by a 2015 prediction.
YarcData has a $100,000 Graphics Analytics Challenge to encourage a move in the industry, with a first prize of $70,000 and several other small prizes. According to White, this challenge is set to “really encourage the development with SPARQL Queries.”
In a couple of weeks, there will be a conference in New York on the semantic web, and YarcData will be sponsoring it with IBM.
To listen to the full interview, you can go here.
