diff --git a/_data/preslist.yml b/_data/preslist.yml index f956dfc..42c5cf9 100644 --- a/_data/preslist.yml +++ b/_data/preslist.yml @@ -1,3 +1,31 @@ +- title: "Advanced optimizations for source transformation based + automatic differentiation" + description: | + Clad is a LLVM/Clang plugin designed to provide automatic differentiation (AD) + for C++ mathematical functions. It generates code for computing derivatives modifying + abstract syntax tree using LLVM compiler features. Clad supports forward- and + reverse-mode differentiation that are effectively used to integrate all kinds of + functions. The typical AD approach in Machine Learning tools records and flattens the + compute graph at runtime, whereas Clad can perform more advanced optimizations at + compile time using a rich program representation provided by the Clang AST. These + optimizations investigate which parts of the computation graph are relevant to + the AD rules. + + One such technique is the “To-Be-Recorded” optimization, which reduces + the memory pressure to the clad tape data structure in the adjoint mode. Another + optimization technique is activity analysis, which discards all derivative + statements that are not relevant to the generated code. In the talk we will explain + compiler-level optimizations specific to AD, and will show some specific examples + of how these analyses have impacted clad applications. + + location: "[MODE 2024](https://indico.cern.ch/event/1380163/)" + date: 2024-09-25 + speaker: Maksym Andriichuk + id: "VVMODE2024" + artifacts: | + [Link to Slides](/assets/presentations/Maksym_Andriichuk_MODE2024_Optimizations.pdf) + highlight: 1 + - title: "Improving BioDynamo's Performance using ROOT C++ Modules" description: | Poster presented at the FOURTH Mode Workshop on Differentiable Programming for Experiment Design diff --git a/assets/presentations/Maksym_Andriichuk_MODE2024_Optimizations.pdf b/assets/presentations/Maksym_Andriichuk_MODE2024_Optimizations.pdf new file mode 100644 index 0000000..ed261b8 Binary files /dev/null and b/assets/presentations/Maksym_Andriichuk_MODE2024_Optimizations.pdf differ