Address comments regarding 05-parallelism-practice

Author: aobolensk

05-parallelism-practice/05-parallelism-practice.tex

@@ -109,14 +109,15 @@ \section{Tasks}
     \item Matrix Multiplication
     \item Systems of Linear Algebraic Equations
     \item Sort
-    \item Computer Graphics
+    \item Image Processing
   \end{itemize}
 \end{frame}
 
 \begin{frame}{Disclaimers}
   \begin{itemize}
     \item All matrices should be stored in linear arrays (not \texttt{std::vector<std::vector<int> >})
     \item Performance should be meauserd on big matrices/vectors
+    \item Total execution time (per test) is more than 1 second
     \item Functionality should be preserved for wide range of processes count
   \end{itemize}
 \end{frame}
@@ -189,7 +190,8 @@ \section{Data Transfer Methods}
   \begin{itemize}
     \item Requirement: Put the task you have chosen in the description
     \item Reference implementation: original MPI function
-    \item Tasks size should be big. Broadcast should send more that one element (vector).
+    \item Tasks size should be big. Broadcast should send more that one element (vector)
+    \item Consider using binary trees to distribute data across different processes
   \end{itemize}
 \end{frame}
 
@@ -291,9 +293,9 @@ \section{Sort}
   \end{itemize}
 \end{frame}
 
-\section{Computer Graphics}
+\section{Image Processing}
 
-\begin{frame}{Computer Graphics}
+\begin{frame}{Image Processing}
   \begin{itemize}
     \item Image Smoothing
     \item Contrast Enhancement