Kitware: Qt and Scientific Computing
Upcoming SlideShare
Loading in...5
×
 

Like this? Share it with your network

Share

Kitware: Qt and Scientific Computing

on

  • 4,875 views

Kitware is involved with Qt in two main areas. First, Kitware is the provider of the CMake (www.cmake.org) build tool which is used to build KDE, the flagship open source Qt project. In addition, ...

Kitware is involved with Qt in two main areas. First, Kitware is the provider of the CMake (www.cmake.org) build tool which is used to build KDE, the flagship open source Qt project. In addition, Kitware has been using Qt for the ParaView (www.paraview.org) project. ParaView is an open-source, multi-platform data analysis and visualization application. With the recent change in Qt licensing Kitware is also starting to use Qt for many other open source scientific and medical projects. For cross platform C++ GUI development Qt now stands alone as really the only viable choice. This talk will cover the CMake build system, and the impact Qt and Qt derived projects are having on state of the art scientific computation.

Presentation by BIll Hoffman held during Qt Developer Days 2009.

http://qt.nokia.com/developer/learning/elearning

Statistics

Views

Total Views
4,875
Views on SlideShare
4,850
Embed Views
25

Actions

Likes
2
Downloads
75
Comments
0

1 Embed 25

http://www.slideshare.net 25

Accessibility

Categories

Upload Details

Uploaded via as Adobe PDF

Usage Rights

CC Attribution-ShareAlike LicenseCC Attribution-ShareAlike License

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Processing…
Post Comment
Edit your comment

Kitware: Qt and Scientific Computing Presentation Transcript

  • 1. Kitware Scientific Computing with Qt 11/16/2009 Bill Hoffman bill.hoffman@kitware.com
  • 2. Overview • Kitware – History and Business Model • Qt at Kitware – past, present and Future • CMake/CTest/CDash 1
  • 3. Kitware: the Company • Founded in 1998 • Founders: 5 previous employees of GE Corporate Research • Privately held, profitable from creation, no debt – Revenues projected at $9 million in 2009 • ~$14 million if subcontractors included – Principally consulting/grants, with support product revenue • Approximately 65 employees; growing rapidly (30% in 2009) – > 25 PhD
  • 4. Business Model • Open source software – Services and support – Consulting – Collaborative R&D • Commercial products – Value-added products – Applications built on high quality, open source base – Custom (proprietary) software frameworks
  • 5. Kitware: Core Technologies Supercomputing Medical Visualization Imaging • Large data • Registration • Parallel computing • Segmentation • Client / Server • Image Processing • Web / grid architectures • Measurement • Human/Computer • Responsive GUI Interaction Software Process Open-Source Computer Toolkits Vision • Insight ToolKit (ITK) Expertise in: • VisualizationToolkit (VTK) • Behavior/event recognition • CMake • Detection and tracking • ParaView • Segmentation • Publications and consulting • Change Detection
  • 6. Supercomputing Visualization • Transform complex, large data into visual representations that are easier to understand and interact with
  • 7. Supercomputing Visualization (2) • Informatics
  • 8. Supercomputing 3D Interaction Widgets
  • 9. Supercomputing Visualization • Applicable software – VTK (the Visualization Toolkit) – ParaView (Parallel Visualization application) – Overview (Information visualization application) • Customers – Any producer and/or consumer of large, complex data
  • 10. Medical Imaging • Process, analyze and visualize medical data – Image processing- operate on images to extract information – Segmentation- identify features in an image – Registration- align images
  • 11. Medical Imaging • Applicable Software – ITK (Insight Segmentation and Registration Toolkit) – IGSTK (Image Guided Surgery Toolkit) – VolView (volume rendering application) – Maverick (medical image analysis application framework) • Customers – Medical research – Pharma – Orthopedics – Medical device manufactures utilizing images – Pre-clinical – Biological applications (image based)
  • 12. Computer Vision • Extract information from images or video streams
  • 13. VTK
  • 14. National Library of Medicine Segmentation and Registration Toolkit $13 million over 6 years Leading edge algorithms Open Source Software www.itk.org
  • 15. ParaView • Parallel Visualization application • Turn-key wrapper around VTK • Supports parallel data processing and rendering • Supports tiled displays, Caves, etc. • 3D widgets, LOD (level-of-detail) display • Extended by XML modules • Extensive animation support
  • 16. Kitware GUI History • 1999 – 2009 C++ GUI Choices – tcl/TK – FLTK -> LGPL – Qt -> GPL – GTK – WxWidgets -> BSD (1992) – MFC (Only Windows) • Requirements – Cross Platform, Open Source “commercializable“ • Kitware created KWWidgets on top of tcl/TK
  • 17. Kitware Qt GPL projects • ParaQ ParaView (converted from KWWidgets) 2005 – Pay to keep BSD license on ParaView/VTK • Maverick DOD SBIR • One commercial customer 16
  • 18. Slider - image-to-image registration app 17
  • 19. Qt LGPL Changes Everything! • Scientists and Engineers can focus on the science and not on GUI widgets and license issues. • Kitware plans to use and encourage the use of Qt for all of our projects now.
  • 20. 3D Slicer • Is a multi-platform, open source software for bio-medical computing • Initially implemented using KWWidgets due to GPL licensing concerns • Utilizes open-source tools CMake, VTK, ITK • $350K of funding for Qt port – NAMIC National Alliance of Medical Image Computing (na- mic.org) – One of seven prestigiuos NIH National Roadmap Projects, a National Center of Biomedical Computing(NCBC) • Two year effort, effort is underway • Cited reasons: Ease of use, cross-platform support, available community resources, and productivity gains
  • 21. CMake – Cross Platform Build System • Why CMake? • CMake Features • Where to get help – Mastering CMake Book – Web Page: www.cmake.org • http://www.cmake.org/Wiki/CMake • mailing list: cmake@cmake.org 21
  • 22. Why CMake? It’s easy, and works well Typical Project without CMake (curl) • A build system $ ls CHANGES RELEASE-NOTES curl-config.in missing CMake acinclude.m4 curl-style.el mkinstalldirs that just works CMakeLists.txt build docs aclocal.m4 depcomp notes~ notes COPYING buildconf include packages CVS buildconf.bat install-sh reconf ChangeLog compile lib sample.emacs Makefile config.guess libcurl.pc.in src Makefile.am config.sub ltmain.sh tests Makefile.in configure m4 vc6curl.dsw README configure.ac maketgz • A build system $ ls src/ CMakeLists.txt Makefile.riscos curlsrc.dsp hugehelp.h version.h CVS Makefile.vc6 curlsrc.dsw macos writeenv.c that is easy to use Makefile.Watcom Makefile.vc8 Makefile.am curlutil.c main.c writeenv.h config-amigaos.h curlutil.h makefile.amiga writeout.c Makefile.b32 config-mac.h getpass.c makefile.dj writeout.h cross platform Makefile.in Makefile.inc config-riscos.h getpass.h mkhelp.pl config-win32.h homedir.c setup.h Makefile.m32 config.h.in homedir.h urlglob.c Makefile.netware curl.rc hugehelp.c urlglob.h 22
  • 23. Why CMake? It’s fast http://blog.qgis.org/?q=node/16 : “I was quite surprised with the speed of building Quantum GIS codebase in comparison to Autotools. “
  • 24. Why CMake? Everyone is using it KDE 2006 – Tipping Point! • 800+ downloads per day from www.cmake.org • Major Linux distributions and Cygwin provide CMake packages • KDE, Second Life, Boost (Expermentally), many others
  • 25. What Is CMake? • Family of Software Development Tools – Build – CMake – Test – CTest/CDash – Package – CPack • Open-Source License • History – Insight Segmentation & Registration Toolkit (~2000) – Changed the way we build
  • 26. How CMake Changes The Way We Build C++ • Boost aims to give C++ a set of useful libraries like Java, Python, and C# • CMake aims to give C++ compile portability like the compile once and run everywhere of Java, Python, and C# – Same build tool and files for all platforms – Easy to mix both large and small libraries
  • 27. Who Is Involved? Users Supporters • KDE • Kitware • Second Life • ITK • ARL • VTK • National Library of • ParaView Medicine • Trilinos • Scribus • Sandia National Labs • Boost (Experimentaly) • Los Alamos National • Mysql Labs • many more • NAMIC 27
  • 28. CMake Features • One simple language for all platforms • Built-in rules for common targets – Windows, Mac, Linux, UNIX variants – Executables – Embedded platforms via cross-compilation – Shared Libraries / DLLs • Generates native build systems – Static Libraries (archives) – Makefiles (GNU, NMake, Borland, etc.) – OS X Frameworks and App Bundles – KDevelop, Eclipse – Visual Studio 6,7,8,9 IDE • Custom rules for other targets – Xcode – Generated Documentation • Out-of-source build trees leave source – Generated sources and headers clean • Configuration rules • Interactive configuration via GUI – System introspection • Multiple configurations (Debug, Release, – Persistent variables (options, etc.) cached results) – Configured header files
  • 29. CMake Features - continued • Automatic analysis – Implicit dependencies (C, C++, Fortran) – Transitive link dependencies – Ordering of linker search path and RPATH • Advanced Makefile generation – Modular, Fast, Parallel – Color and progress display – Help targets – make help – Preprocessor targets – make foo.i – Assembly targets – make foo.s
  • 30. Automatic Testing Benefits “Automated Software Testing,” 1999, Dustin, et al, Addison Wesley
  • 31. CDash Dashboard www.cdash.org
  • 32. Coverage Display GCov/Bullseye
  • 33. Video of ParaView Nightly Testing 33
  • 34. Valgrind / Purify
  • 35. Kitware Hosted CDash http://my.cdash.org
  • 36. CPack • CPack is bundled with CMake • Creates professional platform specific installers – TGZ and Self extract TGZ (STGZ), NullSoft Scriptable Install System (NSIS), OSX PackageMaker, RPM, Deb
  • 37. Thank you