1 / 14

Visual Analytics: An opportunity for the HPC community

Visual Analytics: An opportunity for the HPC community. Shawn J. Bohn Shawn.Bohn@pnl.gov September 8-10, 2008 HPC User Forum Meeting. Visual Analytics. Definitions, What and Why Visual Analytics History of Science Leading into the Future New Requirements within Digital Universe

vaughan-johns
Download Presentation

Visual Analytics: An opportunity for the HPC community

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Visual Analytics:An opportunity for the HPC community Shawn J. Bohn Shawn.Bohn@pnl.gov September 8-10, 2008 HPC User Forum Meeting

  2. Visual Analytics Definitions, What and Why Visual Analytics History of Science Leading into the Future New Requirements within Digital Universe Examples and Observations Conclusion 2

  3. Visual Analytics Definition Congress: Visual analytics provides the last 12 inches between the masses of information and the human mind to make decisions Science: Visual analytics is the science of analytical reasoning facilitated by interactive visual interfaces 3

  4. History of Graphics and Visualization 70s to 80s CAD/CAM Manufacturing, cars, planes, and chips 3D, education, animation, medicine, etc. • 90s to 2000s • Information visualization • Web and Virtual environments • 80s to 90s • Scientific visualization • Realism, entertainment • 2000s to 2010s • Visual Analytics • Visual/audio appliances 4

  5. Selected Societal Drivers and Observations Scale of Things to Come: Information: In 2002, recorded media and electronic information flows generated about 22 exabytes (1018) of information In 2006, the amount of digital information created, captured, and replicated was 161 EB In 2010, the amount of information added annually to the digital universe will be about 988 EB (almost 1 ZB) REFERENCES: A Forecast of Worldwide Information Growth Through 2010: IDC National Open Source Enterprise - Intelligence Community Directive No. 301, July 11, 2006 UC Berkeley School of Information Management and Systems: Now much Information 5

  6. Why Must We Adapt Scale of Things to Come: Information: Drivers of Digital Universe: 70% of the Universe is being produced by individuals Organizations (businesses, agencies, governments, universities) produce 30% Wal-Mart has a database of 0.5 PB; it captures 30,000,000 transactions/day The growth is uneven Today the United States accounts for 41% of the Universe; by 2010, the Asia Pacific region will be growing 40% faster than any of the other regions 6

  7. Why Must We Adapt Scale of Things to Come: Information Drivers of Digital Universe Kinds of Data: About 2 GB of digital information is being produced per person per year 95% of the Digital Universe’s information is unstructured 25% of the digital information produced by 2010 will be images By 2010, the number of e-mailboxes will reach 2 billion The users will send 28 trillion e-mails/year, totaling about 6 EB of data 7

  8. Why Must We Adapt Scale of Things to Come: Information: Drivers of Digital Universe: Kinds of Data Interaction: Today's interaction designed for point and click on individual items, directories, folders, and lists Today's interaction assumes user knows subject, concepts within information spaces, and can articulate what they want Today's interaction assumes data and interconnecting relationships are static in meaning over time Today's interaction is one way initiated Today’s interaction (WIMP) designed over 30 years ago 8

  9. Visual Analytic Engages Multiple Specialties

  10. Examples Demonstrating Need Changing Nature of Information Structure: Temporal, dynamically changing relationships, determination of intent (DC Sniper & ThemeRiver) 10

  11. Examples Demonstrating Need Information synthesis while preserving security and privacy Data signatures that are semantic and scale Firm 5 Firm 6 Firm 4 Firm 7 Firm 3 Country A Firm 8 Firm 2 Firm 9 Firm 1 Firm 10 A Bank Video Images Financial Audio Discover what is there AND discover what isn’t there 11

  12. HPC and Visual Analytics – Example • State-of-the-art1 • Distributed in both processing and data • Scalable (for a single data type) • Limitations • Inability to reconfigure on the fly • Data model is based on subsetting • Batch vs. interactive Krishnan M, SJ Bohn, WE Cowley, VL Crow, and J Nieplocha. 2007. "Scalable Visual Analytics of Massive Textual Datasets." In IPDPS 2007. IEEE International Parallel and Distributed Processing Symposium, 26-30 March 2007, Long Beach, CA, USA. 12

  13. Some Observations End Users want to: Steer their data (INTERACTION) Include and discard data on-the-fly Via Scenarios/Hypothesis generation Surround the user (i.e., data access) Be data scale and modalities agnostic Work from their desktops/mobile device Web/thin client/server type applications No specialized graphics hardware (simple visuals) Customize visualization based on who they are. Developers of Visual Analytics want: Better toolkits and APIs (e.g., Global Arrays) (who should do this?)

  14. Conclusions Visual Analytics is one of the fastest growing fields of study and practice. Practice of interdisciplinary science is required Broadly applies to many aspects of society Visual Analytics is an HPC opportunity Thank you 14

More Related