Optimizing Without Breaking a Sweat

This article describes novel techniques developed to optimize DreamWork Animation's rendering, animation, and special effects applications without recompiling or relinking by preloading highly optimized libraries at run-time.
  • 开发人员
  • 英特尔® C++ Composer XE
  • 英特尔® 集成性能原件
  • Intel® Threading Building Blocks
  • 英特尔® Integrated Native Developer Experience
  • 英特尔® VTune™ 放大器 XE
  • libm
  • libimf
  • malloc
  • visual computing
  • zlib
  • dreamworks
  • memcpy
  • 游戏开发
  • 图形
  • 优化
  • Putting -lm Before User Objects/Libraries on Link Line Can Impact Performance

    Recommended linking model: icc/icpc/ifort [user objs] [user libs] [sys libs] Using -lm (the GNU math library) prior to user-created objects or libraries causes the GNU libm to be used instead of the Intel math library, impacting performance.
  • Linux*
  • C/C++
  • Fortran
  • 英特尔® C++ 编译器
  • Intel® Fortran Compiler
  • math
  • library
  • performance
  • static
  • libm
  • libimf
  • GNU
  • -lm
  • 优化
  • 订阅 libm