\newlabel{ch13:fig:02}{{12.2}{288}}
\newlabel{ch13:list:01}{{12.1}{288}}
\@writefile{lol}{\contentsline {lstlisting}{\numberline {12.1}Skeleton codes of a GPU kernel and a CPU function}{288}}
+\newlabel{ch13:eq:17}{{12.18}{289}}
\@writefile{lof}{\contentsline {figure}{\numberline {12.3}{\ignorespaces Matrix constant coefficients in a three-dimensional domain.\relax }}{290}}
\newlabel{ch13:fig:03}{{12.3}{290}}
-\newlabel{ch13:eq:17}{{12.18}{290}}
\newlabel{ch13:list:02}{{12.2}{290}}
\@writefile{lol}{\contentsline {lstlisting}{\numberline {12.2}GPU kernels of the projected Richardson method}{290}}
-\@writefile{lof}{\contentsline {figure}{\numberline {12.4}{\ignorespaces Computation of a vector element with the projected Richardson method.\relax }}{292}}
-\newlabel{ch13:fig:04}{{12.4}{292}}
+\@writefile{lof}{\contentsline {figure}{\numberline {12.4}{\ignorespaces Computation of a vector element with the projected Richardson method.\relax }}{291}}
+\newlabel{ch13:fig:04}{{12.4}{291}}
\newlabel{ch13:list:03}{{12.3}{292}}
\@writefile{lol}{\contentsline {lstlisting}{\numberline {12.3}Memory access to the cache texture memory}{292}}
\@writefile{toc}{\contentsline {section}{\numberline {12.5}Experimental tests on a GPU cluster}{293}}
\newlabel{ch13:sec:05}{{12.5}{293}}
-\@writefile{lof}{\contentsline {figure}{\numberline {12.5}{\ignorespaces GPU cluster of tests composed of 12 computing nodes (six machines, each with two GPUs.\relax }}{295}}
-\newlabel{ch13:fig:05}{{12.5}{295}}
+\@writefile{lof}{\contentsline {figure}{\numberline {12.5}{\ignorespaces GPU cluster of tests composed of 12 computing nodes (six machines, each with two GPUs.\relax }}{294}}
+\newlabel{ch13:fig:05}{{12.5}{294}}
\@writefile{lot}{\contentsline {table}{\numberline {12.1}{\ignorespaces Execution times in seconds of the parallel projected Richardson method implemented on a cluster of 24 CPU cores.\relax }}{295}}
\newlabel{ch13:tab:01}{{12.1}{295}}
-\@writefile{lot}{\contentsline {table}{\numberline {12.2}{\ignorespaces Execution times in seconds of the parallel projected Richardson method implemented on a cluster of 12 GPUs.\relax }}{296}}
-\newlabel{ch13:tab:02}{{12.2}{296}}
+\@writefile{lot}{\contentsline {table}{\numberline {12.2}{\ignorespaces Execution times in seconds of the parallel projected Richardson method implemented on a cluster of 12 GPUs.\relax }}{295}}
+\newlabel{ch13:tab:02}{{12.2}{295}}
\@writefile{toc}{\contentsline {section}{\numberline {12.6}Red-Black ordering technique}{296}}
\newlabel{ch13:sec:06}{{12.6}{296}}
+\newlabel{ch13:fig:06.01}{{12.6(a)}{297}}
+\newlabel{sub@ch13:fig:06.01}{{(a)}{297}}
+\newlabel{ch13:fig:06.02}{{12.6(b)}{297}}
+\newlabel{sub@ch13:fig:06.02}{{(b)}{297}}
+\@writefile{lof}{\contentsline {figure}{\numberline {12.6}{\ignorespaces Red-Black ordering for computing the iterate vector elements in a three-dimensional space.\relax }}{297}}
+\@writefile{lof}{\contentsline {subfigure}{\numberline{(a)}{\ignorespaces {Red-Black ordering on x, y and z axises}}}{297}}
+\@writefile{lof}{\contentsline {subfigure}{\numberline{(b)}{\ignorespaces {Red-Black ordering on y axis}}}{297}}
\newlabel{ch13:list:04}{{12.4}{297}}
\@writefile{lol}{\contentsline {lstlisting}{\numberline {12.4}GPU kernels of the projected Richardson method using the red-black technique}{297}}
-\newlabel{ch13:fig:06.01}{{12.6(a)}{298}}
-\newlabel{sub@ch13:fig:06.01}{{(a)}{298}}
-\newlabel{ch13:fig:06.02}{{12.6(b)}{298}}
-\newlabel{sub@ch13:fig:06.02}{{(b)}{298}}
-\@writefile{lof}{\contentsline {figure}{\numberline {12.6}{\ignorespaces Red-Black ordering for computing the iterate vector elements in a three-dimensional space.\relax }}{298}}
-\@writefile{lof}{\contentsline {subfigure}{\numberline{(a)}{\ignorespaces {Red-Black ordering on x, y and z axises}}}{298}}
-\@writefile{lof}{\contentsline {subfigure}{\numberline{(b)}{\ignorespaces {Red-Black ordering on y axis}}}{298}}
\@writefile{lot}{\contentsline {table}{\numberline {12.3}{\ignorespaces Execution times in seconds of the parallel projected Richardson method using read-black ordering technique implemented on a cluster of 12 GPUs.\relax }}{299}}
\newlabel{ch13:tab:03}{{12.3}{299}}
\@writefile{lof}{\contentsline {figure}{\numberline {12.7}{\ignorespaces Weak scaling of both synchronous and asynchronous algorithms of the projected Richardson method using red-black ordering technique.\relax }}{300}}
