28 lines
1017 B
HTML
28 lines
1017 B
HTML
|
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
|
||
|
<html>
|
||
|
<head>
|
||
|
<link rel="stylesheet" type="text/css" href="css.css">
|
||
|
</head>
|
||
|
<body>
|
||
|
<table class="doctable" border="0" cellspacing="0" summary=
|
||
|
"this table is just for layout">
|
||
|
<tr class="toprow"><td class=maintitle>the vx-scheme compiler
|
||
|
<p align="right">[ <a href="download.html">Download</a> ]</p>
|
||
|
</td></tr>
|
||
|
<tr><td class=body>
|
||
|
|
||
|
<p><img align="right" src="benchmark.png"> This was interesting for
|
||
|
me: it's some benchmark data collected over all versions of this
|
||
|
program starting from version 0.4 and up to the compiler in version
|
||
|
0.7. Most of the effort in tuning the program has paid off. The
|
||
|
compilation occasionally loses vs. interpretation in very short
|
||
|
programs or those that don't loop much (you can see this effect
|
||
|
in <code>dynamic</code> and <code>dderiv</code>. </p>
|
||
|
<p>Otherwise, in die-hard computational scenarios
|
||
|
like <code>ack</code> and <code>puzzle</code>, the compiler is seen at
|
||
|
its best.</p>
|
||
|
</td></tr>
|
||
|
</table>
|
||
|
</body>
|
||
|
</html>
|