<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
<tt>Fyi,<br>
Double Intel Xeon 3Ghz performance below<br>
<br>
g723 gsm ulaw alaw g726 adpcm slin lpc10 g729 speex
ilbc<br>
g723 - - - - - - - - - -
-<br>
gsm - - 2 2 2 2 1 4 10 29
14<br>
ulaw - 2 - 1 2 2 1 4 10 29
14<br>
alaw - 2 1 - 2 2 1 4 10 29
14<br>
g726 - 2 2 2 - 2 1 4 10 29
14<br>
adpcm - 2 2 2 2 - 1 4 10 29
14<br>
slin - 1 1 1 1 1 - 3 9 28
13<br>
lpc10 - 3 3 3 3 3 2 - 11 30
15<br>
g729 - 3 3 3 3 3 2 5 - 30
15<br>
speex - 3 3 3 3 3 2 5 11 -
15<br>
ilbc - 3 3 3 3 3 2 5 11 30
-<br>
<br>
Olivier<br>
</tt><br>
<br>
Tzafrir Cohen a écrit :
<blockquote cite="mid20060709122042.GS12623@xorcom.com" type="cite">
<pre wrap="">On Sun, Jul 09, 2006 at 05:07:16AM -0400, C F wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Tzafrir, are you trying to tell me that I can realy do double on the
intel becuase the second CPU will do it?
</pre>
</blockquote>
<pre wrap=""><!---->
In the ideal case you'll get double performance with two CPUs. In
theory.
A case of many concurrent calls is basically something that can be
easily parallelized. So in theory nothing stops you from getting
something closer to double performance. I don't know how close reality
is to that nice theory.
I only remarked that 'show translations' totally ignores the second CPU.
</pre>
</blockquote>
</body>
</html>