11 4 VisIt is a richly featured, turnkey application VisIt is an open source, end user visualization and analysis tool for simulated and experimental data –Used by: physicists, engineers, code developers, vis experts –>100K downloads on web R&D 100 award in 2005 Used “heavily to exclusively” on 8 of world’s top 12 supercomputers 217 pin reactor cooling simulation. Run on ¼ of Argonne BG/P. 1 billion grid points

22 Terribly Named!! Intended for more than just visualization ! Data Exploration Visual Debugging Quantitative Analysis Presentations = ? Comparative Analysis

33 VisIt has a rich feature set that can impact many science areas.  Meshes: rectilinear, curvilinear, unstructured, point, AMR  Data: scalar, vector, tensor, material, species  Dimension: 1D, 2D, 3D, time varying  Rendering (~15): pseudocolor, volume rendering, hedgehogs, glyphs, mesh lines, etc…  Data manipulation (~40): slicing, contouring, clipping, thresholding, restrict to box, reflect, project, revolve, …  File formats (~85)  Derived quantities: >100 interoperable building blocks +,-,*,/, gradient, mesh quality, if-then-else, and, or, not  Many general features: position lights, make movie, etc  Queries (~50): ways to pull out quantitative information, debugging, comparative analysis

44 VisIt employs a parallelized client-server architecture.  Client-server observations: Good for remote visualization Leverages available resources Scales well No need to move data  Additional design considerations: Plugins Multiple UIs: GUI (Qt), CLI (Python), more… remote machine Parallel vis resources User data localhost – Linux, Windows, Mac Graphics Hardware

55 The VisIt team focuses on making a robust, usable product for end users.  Manuals 300 page user manual 200 page command line interface manual “Getting your data into VisIt” manual  Wiki for users (and developers)  Revision control, nightly regression testing, etc  Executables for all major platforms  Day long class, complete with exercises Slides from the VisIt class

66 VisIt is a vibrant project with many participants.  Over 50 person-years of effort  Over one million lines of code  Partnership between: Department of Energy’s Office of Nuclear Energy, Office of Science, and National Nuclear Security Agency, and among others User community grows, including AWE & ASC Alliance schools Fall ‘06 VACET is funded Spring ‘08 AWE enters repo 2003 LLNL user community transitioned to VisIt R&D SciDAC Outreach Center enables Public SW repo 2007 Saudi Aramco funds LLNL to support VisIt Spring ‘07 GNEP funds LLNL to support GNEP codes at Argonne Summer‘07 Developers from LLNL, LBL, & ORNL Start dev in repo ‘07-’08 UC Davis & UUtah research done in VisIt repo 2000 Project started ‘07-’08 Partnership with CEA is developed 2008 Institutional support leverages effort from many labs Spring ‘09 More developers Entering repo all the time

7 VisIt: What’s the Big Deal?  Everything works at scale  Robust, usable tool  Vis to code development to scientific insight 7

8 ● Weak scaling study: ~62.5M cells/core 8 #coresProblem Size ModelMachine 8K0.5TIBM P5Purple 16K1TSunRanger 16K1TX86_64Juno 32K2TCray XT5JaguarPF 64K4TBG/PDawn 16K, 32K1T, 2TCray XT4Franklin VisIt was recently demonstrated to show good performance at unprecedented scales.

9 Three Ways To Get Data Into VisIt  (1) Write to a known output format  (2) Write a plugin file format reader  (3) Integrate VisIt “in situ” “lib-VisIt” is linked into simulation code −(Note: Memory footprint issues with implementation!) Use model: −simulation code advances −at some time interval (e.g. end of cycle), hands control to lib-VisIt. −lib-VisIt performs vis & analysis tasks, then hands control back to simulation code −repeat 9

10 Tutorial schedule  10 minute overview (done)  6 part tutorial on using VisIt Introduction Data analysis / expressions Scripting Moviemaking Comparisons Alternate data representations  Data model overview  Database plugin development example  Operator plugin development example  Presentation on VisIt architecture / parallel algorithms 10

11 Before we begin…  Tutorial: Hank Childs ( ), Sean Ahern ( ) Script: Another tutorial upcoming SC09  User resources: Main website: Wiki:  Development resources: SVN: 11

12 AFTER THIS POINT IS THE SLIDES TO CLOSE THE TUTORIAL 12

Outline  Parallelism Contracts Volume rendering Streamlines  Data model AMR Subset Inclusion Lattices CSG  Miscellaneous More on steering Synthetic Diagnostics 13

14 Contracts are an extension to the standard data flow network design.  Work is performed by a pipeline  A pipeline consists of data objects and components (sources, filters, and sinks) File Reader (Source) Slice Filter Contour Filter Renderer (Sink) V0V1 V2Exec Update Execute Data Flow Networks “101”: Extension: Contracts are coupled with the Update phase Pipeline execution begins with a “pull”, which starts Update phase Data flows from component to component during the Execute phase

15 Operating on Optimal Subset of Data … Slice Filter V( I ) … V( I+1 ) Filter (base class) (inheritance) Slice Filter Hundreds of others Contract * SliceFilter::ModifyContract (Contract *V(I)) {... return V(I+1); } 1)Get meta-data 2)Determine domains that intersect slice 3)Restrict list of domains to process in V( I +1)

16 Operating on Optimal Subset of Data … Slice Filter V( I ) … V( I+1 ) 1)Get meta-data 2)Determine domains that intersect slice 3)Restrict list of domains to process in V( I +1) D3D2D1D0 D3D2D1D0

17 The contract-based system provides high flexibility for this optimization. … Spherical Slice Filter V( I ) … V( I+1 ) A new, plugin filter can use this optimization without any modification to base data flow network architecture … Slice Filter V( J ) … V( J+1 ) … Contour Filter V( I+1 ) V( I ) … … Multiple filters can use the same optimizations

Visualization of massive data requires sophisticated algorithms. “Image Space” Partition Redundant I/O Each frame requires new partition “World Space” Partition Camera inside data set causes huge imbalance. A hybrid approach that acts as a world space partition, but identifies regions of imbalance and handles those regions using an image space partition.

This analysis depended on improvements in parallel particle advection  Two extremes: Partition data over processors and pass particles amongst processors −Parallel inefficiency! Partition seed points over processors and process necessary data for advection −Redundant I/O! Notional streamline example Notional streamline example P0 P1 P2 P3 P4 P0 P1 P2 P3 P4

 Hybrid solution: Master-slave approach that adapts between parallel inefficiencies and redundant I/O P0 P1 P2 P3 P4 IterationAction 0P0 reads B0, P3 reads B1 1P1 passes points to P0, P4 passes points to P3, P2 reads B0 0: Read Notional streamline example Notional streamline example 1: Pass 1: Read - Decision of when to pass or read is simplified here and based on load of processors owning data. - Heuristic weights against I/O, but allows for redundant I/O - All coordination done by slave masters. This analysis depended on improvements in parallel particle advection

21 AMR is handled natively in VisIt  VisIt understands: Nesting of patches / Boundaries between patches  Strategy: Each patch is processed separately After reading data, subsequent passes to: −Identify coarse elements that are refined and mark them as “ghost” −Create ghost layers around outer boundaries (needs work)  UI: Color by patch, color by level Remove / show patches, levels Have appropriate info returned for picks

22 CSG Visualization  Adaptively creates a volumetric representation  Supported shapes Sphere Cone Cylinder Plane Quadric Tetrahedron Pyramid Prism Hexahedron

23 Users need to reference their data “as they know it”.  Many different decompositions of data: By simulation processor ID By material By part in an assembly By level/patch for AMR Plus key subsets, nodes/elements representing: −Boundary conditions −Slide surfaces −Probes We need to allow users to cull away unneeded subsets of the data.

24 Representing data relationships  Relationships between many decompositions are complicated.  Need sophisticated machinery to store this knowledge.  Machinery must be customizable to give right “look and feel”.  Idea: incorporate Subset Inclusion Lattices (SIL)

25 Subset Inclusion Lattices (SILs)  SIL: directed, acyclic graph  Bipartite: one set of nodes for subsets, the other for categories  Encodes relationships between sets (i.e. subset relations) Also encodes “category” of that relation (i.e. material, processor ID, etc)

26 VisIt has full built-in support for SILs  VisIt supports over fifty file formats. Now each can present data subsets to users in an appropriate way.  VisIt can build GUI windows, etc, based on a SIL construction.  VisIt creates SILs based on basic information (material names, # of processors, etc) Only have to modify a SIL for advanced cases.

27 ● Simple steering – Scalar inputs to the simulation – Specified through a GUI ● Dashboard of status ● Built by extending our simulation monitoring ● Simulation code can specify a custom GUI window – Loaded through a Qt UI file built with Qt Designer – State is synchronized between VisIt and simulation 27 Simulation name Simulation value pairs Simulation status Engine controls Commands control Component name Text value Simulation steering

28 Quantitative analysis means different things to different people. 1)Techniques that span scientific domains (e.g. integration, volumes, surface areas, fluxes, etc.) 2)Techniques specific to scientific domains (e.g. hohlraum flux at AGEX) Detector at AGEX Detector provided by VisIt (synthetic diagnostic)

29 … and we’re done.  Tutorial: Hank Childs ( ), Sean Ahern ( ) Script: Another tutorial upcoming SC09  User resources: Main website: Wiki:  Development resources: SVN: 29