Fix includegraphics'.

diff --git a/Heter_paper.tex b/Heter_paper.tex
index 3b2b238cbda8df971488379ffe7178421ca188fa..a1f6b2226c061186ff41f7a45b4f5f0884d2708d 100644 (file)
--- a/Heter_paper.tex
+++ b/Heter_paper.tex
@@ -279,10 +279,10 @@ execution time, the normalized performance,  as follows:
 \begin{figure}
   \centering
   \subfloat[Homogeneous platform]{%
 \begin{figure}
   \centering
   \subfloat[Homogeneous platform]{%

-    \includegraphics[width=.22\textwidth]{fig/homo.eps}\label{fig:r1}}%
+    \includegraphics[width=.22\textwidth]{fig/homo}\label{fig:r1}}%

   \qquad%
   \subfloat[Heterogeneous platform]{%
   \qquad%
   \subfloat[Heterogeneous platform]{%

-    \includegraphics[width=.22\textwidth]{fig/heter.eps}\label{fig:r2}}
+    \includegraphics[width=.22\textwidth]{fig/heter}\label{fig:r2}}

   \label{fig:rel}
   \caption{The energy and performance relation}
 \end{figure}
   \label{fig:rel}
   \caption{The energy and performance relation}
 \end{figure}


@@ -312,7 +312,7 @@ In this section we proposed an heterogeneous scaling algorithm, (figure~\ref{HSA
 The algorithm is numerates the suitable range of available scaling factors for each node in the heterogeneous cluster, returns a set of optimal frequency scaling factors for each node. Using heterogeneous cluster is produces different workloads for each node. Therefore, the fastest nodes waiting at the barrier for the slowest nodes to finish there work as in figure (\ref{fig:heter}). Our algorithm takes into account these imbalanced workloads when is starts to search for selecting the best scaling factors. So, the algorithm is selecting the initial frequencies values for each node proportional to the times of computations that gathered from the first iteration. As an example in figure (\ref{fig:st_freq}), the algorithm don't test the first frequencies of the fastest nodes until it converge their frequencies to the frequency of the slowest node. If the algorithm is starts test changing the frequency of the slowest nodes from beginning, we are loosing performance and then not selecting the best tradeoff (the distance). This case will be similar to the homogeneous cluster when all nodes scales their frequencies together from the  beginning. In this case there is a small distance between energy and performance curves, for example see the figure(\ref{fig:r1}).  Then the algorithm searching for optimal frequency  scaling factor from the selected frequencies until the last available ones. 
 \begin{figure}[t]
   \centering
 The algorithm is numerates the suitable range of available scaling factors for each node in the heterogeneous cluster, returns a set of optimal frequency scaling factors for each node. Using heterogeneous cluster is produces different workloads for each node. Therefore, the fastest nodes waiting at the barrier for the slowest nodes to finish there work as in figure (\ref{fig:heter}). Our algorithm takes into account these imbalanced workloads when is starts to search for selecting the best scaling factors. So, the algorithm is selecting the initial frequencies values for each node proportional to the times of computations that gathered from the first iteration. As an example in figure (\ref{fig:st_freq}), the algorithm don't test the first frequencies of the fastest nodes until it converge their frequencies to the frequency of the slowest node. If the algorithm is starts test changing the frequency of the slowest nodes from beginning, we are loosing performance and then not selecting the best tradeoff (the distance). This case will be similar to the homogeneous cluster when all nodes scales their frequencies together from the  beginning. In this case there is a small distance between energy and performance curves, for example see the figure(\ref{fig:r1}).  Then the algorithm searching for optimal frequency  scaling factor from the selected frequencies until the last available ones. 
 \begin{figure}[t]
   \centering

-    \includegraphics[scale=0.5]{fig/start_freq.pdf}
+    \includegraphics[scale=0.5]{fig/start_freq}

   \caption{Selecting the initial frequencies}
   \label{fig:st_freq}
 \end{figure}
   \caption{Selecting the initial frequencies}
   \label{fig:st_freq}
 \end{figure}


@@ -615,10 +615,10 @@ The results of applying the proposed scaling algorithm to NAS benchmarks is demo
 \begin{figure}
   \centering
   \subfloat[Balanced nodes type scenario]{%
 \begin{figure}
   \centering
   \subfloat[Balanced nodes type scenario]{%

-    \includegraphics[width=.23185\textwidth]{fig/avg_eq.eps}\label{fig:avg_eq}}%
+    \includegraphics[width=.23185\textwidth]{fig/avg_eq}\label{fig:avg_eq}}%

   \quad%
   \subfloat[Imbalanced nodes type scenario]{%
   \quad%
   \subfloat[Imbalanced nodes type scenario]{%

-    \includegraphics[width=.23185\textwidth]{fig/avg_neq.eps}\label{fig:avg_neq}}
+    \includegraphics[width=.23185\textwidth]{fig/avg_neq}\label{fig:avg_neq}}

   \label{fig:avg}
   \caption{The average of energy and performance for all Nas benchmarks running with difference number of nodes}
 \end{figure}
   \label{fig:avg}
   \caption{The average of energy and performance for all Nas benchmarks running with difference number of nodes}
 \end{figure}


@@ -688,10 +688,10 @@ The results of the previous section are obtained using a percentage of 80\% for
 \begin{figure}
   \centering
   \subfloat[Comparison the average of the results on 8 nodes]{%
 \begin{figure}
   \centering
   \subfloat[Comparison the average of the results on 8 nodes]{%

-    \includegraphics[width=.22\textwidth]{fig/sen_comp.pdf}\label{fig:sen_comp}}%
+    \includegraphics[width=.22\textwidth]{fig/sen_comp}\label{fig:sen_comp}}%

   \quad%
   \subfloat[Comparison the selected frequency scaling factors for 8 nodes]{%
   \quad%
   \subfloat[Comparison the selected frequency scaling factors for 8 nodes]{%

-    \includegraphics[width=.24\textwidth]{fig/three_scenarios.pdf}\label{fig:scales_comp}}
+    \includegraphics[width=.24\textwidth]{fig/three_scenarios}\label{fig:scales_comp}}

   \label{fig:avg}
   \caption{The comparison of the three power scenarios}
 \end{figure}
   \label{fig:avg}
   \caption{The comparison of the three power scenarios}
 \end{figure}

diff --git a/fig/Makefile b/fig/Makefile
new file mode 100644 (file)
index 0000000..f8cbf58
--- /dev/null
+++ b/fig/Makefile
@@ -0,0 +1,12 @@
+EPS = $(wildcard *.eps)
+PDF = $(EPS:%.eps=%.pdf)
+
+.PHONY: all clean
+
+all: $(PDF)
+
+clean:
+       $(RM) $(PDF)
+
+%.pdf: %.eps
+       epstopdf $<

diff --git a/fig/avg_eq-eps-converted-to.pdf b/fig/avg_eq.pdf
similarity index 94%
rename from fig/avg_eq-eps-converted-to.pdf
rename to fig/avg_eq.pdf
index 39c3fd5022d74720c2d7fe18617321bf58a03123..26c5a4aed4b99e501a20cdc8cea4009148737196 100644 (file)
Binary files a/fig/avg_eq-eps-converted-to.pdf and b/fig/avg_eq.pdf differ

diff --git a/fig/avg_neq-eps-converted-to.pdf b/fig/avg_neq.pdf
similarity index 94%
rename from fig/avg_neq-eps-converted-to.pdf
rename to fig/avg_neq.pdf
index 22b5196f6c1b0168aac104eca3b06a967806b018..028c084208d4eb82806b682e11df46038a380d06 100644 (file)
Binary files a/fig/avg_neq-eps-converted-to.pdf and b/fig/avg_neq.pdf differ

diff --git a/fig/heter-eps-converted-to.pdf b/fig/heter.pdf
similarity index 93%
rename from fig/heter-eps-converted-to.pdf
rename to fig/heter.pdf
index 5cf96d4e6075498d0acd0ceb274610aeacdf7ac9..31e3fda90beffd3cf10ef9bccafc096fe98794a3 100644 (file)
Binary files a/fig/heter-eps-converted-to.pdf and b/fig/heter.pdf differ

diff --git a/fig/homo.eps b/fig/homo.eps
old mode 100755 (executable)
new mode 100644 (file)

diff --git a/fig/homo-eps-converted-to.pdf b/fig/homo.pdf
similarity index 94%
rename from fig/homo-eps-converted-to.pdf
rename to fig/homo.pdf
index a82a1aeea64676a284e4ac456cf8ef6d0e8fd04b..2d4cb51c2213e8f66c25e56daa59b0a3e872fae5 100644 (file)
Binary files a/fig/homo-eps-converted-to.pdf and b/fig/homo.pdf differ