[pypy-svn] r80237 - pypy/extradoc/talk/pepm2011/presentation
cfbolz at codespeak.net
cfbolz at codespeak.net
Mon Jan 24 22:00:45 CET 2011
Author: cfbolz
Date: Mon Jan 24 22:00:43 2011
New Revision: 80237
Modified:
pypy/extradoc/talk/pepm2011/presentation/talk.tex
Log:
last tweaks
Modified: pypy/extradoc/talk/pepm2011/presentation/talk.tex
==============================================================================
--- pypy/extradoc/talk/pepm2011/presentation/talk.tex (original)
+++ pypy/extradoc/talk/pepm2011/presentation/talk.tex Mon Jan 24 22:00:43 2011
@@ -158,7 +158,7 @@
\item \textbf{Add an optimization that can deal with heap operations}
\pause
\item optimize short-lived objects
- \item remove some of the redundancy
+ \item remove some of the redundant type checks
\end{itemize}
\end{frame}
@@ -262,6 +262,7 @@
i6 = int_add(i4, i5)
y = new(Integer)
set(y, intval, i6)
+return(y)
\end{verbatim}
\end{frame}
@@ -285,6 +286,7 @@
i6 = int_add(\emph{i3}, i5)
y = new(Integer)
set(y, intval, i6)
+return(y)
\end{alltt}
\end{frame}
@@ -498,6 +500,24 @@
\includegraphics[scale=0.8]{figures/opt_set_dynamic2}
\end{frame}
+\begin{frame}
+ \frametitle{Properties of the Optimization}
+ \begin{itemize}
+ \item output trace is never longer than input trace
+ \item runtime linear in the length of the trace
+ \end{itemize}
+ \pause
+ \begin{block}{Implementation}
+ \begin{itemize}
+ \item about 400 lines of code
+ \item some added complexity over presentation
+ \item objects with arbitrary numbers of fields
+ \item array support
+ \end{itemize}
+ \end{block}
+\end{frame}
+
+
\begin{frame}[plain]
\frametitle{Optimizing the Example Trace}
\only<1>
@@ -542,20 +562,22 @@
\frametitle{Benchmark Results}
\begin{itemize}
\item to evaluate the optimization we used PyPy's Python interpreter with real-world programs
+ \begin{itemize}
+ \item interpreter is about 30'000 lines of code
+ \end{itemize}
+ \pause
\item optimization can remove
\begin{itemize}
\item 70\% of all \texttt{new} operations
\item 14\% of all \texttt{get/set} operations
\item 93\% of all \texttt{guard} operations
+ \end{itemize}
\pause
\item Timings improve by a factor between 1.1 and 6.95
\item outperforming standard Python on all benchmarks but two
- \end{itemize}
\end{itemize}
\end{frame}
-\section{Benchmarks}
-
\begin{frame}
\frametitle{Benchmark}
\includegraphics[scale=0.4]{figures/benchmarks}
More information about the Pypy-commit
mailing list