It won’t be long before business managers, researchers, analysts, professors and their students pick up their smartphones and, using the nearest wall for a monitor screen, start computing with a distant facility too powerful for any but the wealthiest organizations to maintain by themselves. That’s the promise of cloud computing heard by attendees at the CIMS Sponsors Meeting in October.
“Ask a hundred people what cloud computing is, and you will get 100 different answers,” NetApp chief executive Tom Georges told Barron’s magazine in November. Notwithstanding, Mladen A.Vouk, NC State professor and head of computer science, defines cloud computing as “ a seamless component-based architecture that can deliver an integrated, orchestrated and rich suite of both loosely and tightly coupled on-demand information technology functions and services, and significantly reduce overhead and total cost of ownership and services.”
Prof. Vouk foresees an order of magnitude minimum improvement in the cost of computing. This, in turn, will permit raising increasing numbers of people up to performing greater value-added work.
Public Utility in the Clouds
CIMS Industry Fellow Michael Kowolenko sees the cloud as basically a public utility. When combined with the new natural language processing capabilities, it will allow users to plow through the universe of unstructured data efficiently and turn that data into intelligence.
As Prof. Kowolenko wrote in the Fall 2010 Technology Management Report, “We need to develop new tools that can be applied to a wide variety of applications and that many different people can access and use easily.” And that’s what Kowolenko’s Virtual Computing Lab at NC State (http://vcl.ncsu.edu) is doing in its partnership with IBM’s jStart Emerging Software Technologies team (www.ibm.com/jstart)
The VCL describes itself as, “an opensource implementation of a secure production-level technology for widearea access to solutions based on real and virtualized computational, storage, network, and software resources.”
jStart is “a highly specialized group of emerging technology experts within IBM dedicated to leveraging emerging technology to meet real-world IT challenges, providing real returns on technology investments, and helping clients to understand the impact of emerging technologies on their business operations.”
The jStart team identifies four “Big Data” challenges: To harness a tsunami of data in the petabyte range; perform analytics on thousands of transactions a second; correlate structured, semi-structured and unstructured information together; constantly update data analysis and predictive models.
The CIMS-IBM Partnership
jStart’s collaboration with NC State, announced last August, commenced with the use of IBM’s advanced analytics technology to accelerate the lengthy search processes used by the Office of Technology Transfer in finding potential licensees for university biomedical discoveries. (See “Business Intelligence in the Clouds,” on page 1.)
In one case, the months it took dozens of people to search for potential investors and partners for its vaccine research were cut to one week in which IBM Big Data analytics technology was able to analyze 1.4 million Web pages.
At the same time, CIMS and IBM teamed up with the Drug Discovery Center for Innovation to demonstrate the power of the cloud for drug development. The DDCOI (www.ddcoi. org) is a non-profit center designed to serve as a bridge between industry and academia in translating drug discoveries into effective new medicines. It is headquartered in Research Triangle Park, North Carolina, and funded by the NC Biotechnology Center (www. ncbiotech.org).
For its part in the program, the DDCOI is developing and testing a deep search engine dubbed Business Intelligence Gateway, or “BIG.” It would use the VCL’s 2,600-blade server supercomputer to mine both structured and unstructured data sources for new drug discovery opportunities and competitive intelligence information tailored to user needs and objectives.
BIG would be “a pharmaceutical industry Google without the noise,” DDCOI president and CEO John Didsbury told the CIMS Sponsors meeting.
Didsbury reported that the recently completed beta test of the system successfully answered “real life” queries from two large pharmaceutical companies. New competitive intelligence was obtained, as well as a new drug discovery opportunity.
Next steps are to create a business plan for BIG, refine the software tools used and enter a production phase.
Eventually, BIG may be established as a for-profit DDCOI subsidiary and marketed to prospective pharma business partners, Disbury said.
Where We Are
“We’ve now demonstrated that natural language processing algorithms can be applied to business intelligence questions and relevant information extracted from the Internet,” Kowolenko says. “We’re able to separate signal from noise.”
As the program now moves from the proof-of-concept phase to development, “we have to show that we can apply data-driven decision making on a much larger scale,” Kowolenko continues. “Previously, we searched 400 websites and now we need to scan thousands; we’re moving into analyzing terabytes and eventually petabytes.”
All of this should be good news for, among others, the 543 C-level and other business executives in 17 countries who responded to a recent survey conducted for Seattle, Washington-based Avanade by Kelton Research (http://www.avanade. com/BigData). Fifty-six percent of the respondents said they were overwhelmed by the amount of data their company must manage; 62% of the C-level execs complained of being frequently interrupted by irrelevant incoming data; 46% of the companies reported that bad or outdated data had led them to make inaccurate business decisions.
As Rod Smith, IBM vice president of software technology, declared when the CIMS partnership was announced, “The volumes of data on our planet are growing exponentially, which represents huge opportunities for organizations that can unlock the insights hidden within the mountains of information. NC State University sets an example of using smart analysis of big volumes of data to explore and kick start new businesses that push our economy forward.”