https://community.intel.com/t5/Intel-Moderncode-for-Parallel/Halide-MIT-quot-Decoupling-Algorithms-from-Schedules-for-Easy/m-p/768821#M98 <P>Hi. Halide may be of interest to other forums too, like mobile, or Atom, or ...?</P><P>Halide's idea is to separates algorithm from "scheduling"</P><P>The performance numbers are impressive, and probably product differentiators.</P><P>The authors, Ragan-Kelly, Adams, Paris, Levoy, Amarasinghe, and Durand, presented at SIGRAPH 2012.</P><P>If you attended their session I'd appreciate hearing your take.</P><P>Parallel, coarse an fine, seems to, finally, be emerging as more real products are introduced.</P><P>I'm looking forward to Phi.</P> Sat, 11 Aug 2012 16:53:27 GMT Art_Scott 2012-08-11T16:53:27Z https://community.intel.com/t5/Intel-Moderncode-for-Parallel/Halide-MIT-quot-Decoupling-Algorithms-from-Schedules-for-Easy/m-p/768821#M98 <P>Hi. Halide may be of interest to other forums too, like mobile, or Atom, or ...?</P><P>Halide's idea is to separates algorithm from "scheduling"</P><P>The performance numbers are impressive, and probably product differentiators.</P><P>The authors, Ragan-Kelly, Adams, Paris, Levoy, Amarasinghe, and Durand, presented at SIGRAPH 2012.</P><P>If you attended their session I'd appreciate hearing your take.</P><P>Parallel, coarse an fine, seems to, finally, be emerging as more real products are introduced.</P><P>I'm looking forward to Phi.</P> Sat, 11 Aug 2012 16:53:27 GMT https://community.intel.com/t5/Intel-Moderncode-for-Parallel/Halide-MIT-quot-Decoupling-Algorithms-from-Schedules-for-Easy/m-p/768821#M98 Art_Scott 2012-08-11T16:53:27Z