the paper <b>Audio and the Graphics Processing Unit </b>( <a href="http://www.node99.org/projects/gpuaudio/">http://www.node99.org/projects/gpuaudio/</a> ) may be of some usefulness to you, it's not the most positive paper considering what you want to do but eh, it's something...
<a href="http://www.gpgpu.org/">http://www.gpgpu.org/</a> also has a lot of resources that you might want to check out as well.<br><br><br><div><span class="gmail_quote">On 3/4/07, <b class="gmail_sendername">Matthew Rubenstein
</b> <<a href="mailto:email@mattruby.com">email@mattruby.com</a>> wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"> Compression algorithms have generally not been ported to GPUs like the
<br>G80. They're usually more logic and branch oriented than just brute<br>force multiply-accumulates that GPUs specialize in. I also haven't seen<br>any of the popular Asterisk codecs, like G.729 or GSM, ported to any
<br>GPU. Is there a source for codecs ported to GPUs? Or any research<br>showing a good approach?<br><br><br>On Sun, 2007-03-04 at 10:12 -0700, <a href="mailto:asterisk-dev-request@lists.digium.com">asterisk-dev-request@lists.digium.com
</a><br>wrote:<br>> Date: Sat, 03 Mar 2007 22:05:19 -0500<br>> From: Wai Wu <<a href="mailto:wkwu@calltrol.com">wkwu@calltrol.com</a>><br>> Subject: [asterisk-dev] nVidia Cuda<br>> To: <a href="mailto:asterisk-dev@lists.digium.com">
asterisk-dev@lists.digium.com</a><br>> Message-ID: <<a href="mailto:B0430B20D208514CB2AFF57E81645C3101B970@k3-1.Calltrol.com">B0430B20D208514CB2AFF57E81645C3101B970@k3-1.Calltrol.com</a>><br>> Content-Type: text/plain; charset="iso-8859-1"
<br>><br>> Hi devs,<br>><br>> Has any one looked into cuda to see if cpu intensive part of<br>> asterisk(like codecs and conference) can be moved to the G80<br>> processor? I found that the GF8800 cards are very inexpensive (around
<br>> 400 USDs per). I have ported some of our financial applications to<br>> this board and found almost 10x performance improvement over the 3GHz<br>> C2D host processor. I would like to do the same for asterisk.
<br>><br>> To start. I have been looking into the asterisk code and have no crew<br>> how it is structured except the addon applications. Here I have two<br>> questions.<br>><br>> 1) Is each channel use their own thread if a codec is used?
<br>> 2) Which part of the asterisk code actually makes the call to the<br>> necessary codec? I notice in the applications. They save and set the<br>> frame format, then read from the channel. So I trace the read
<br>> function, but all it does is reading from a file descriptor. So I need<br>> some help here.<br>--<br><br>(C) Matthew Rubenstein<br><br>_______________________________________________<br>--Bandwidth and Colocation provided by
<a href="http://Easynews.com">Easynews.com</a> --<br><br>asterisk-dev mailing list<br>To UNSUBSCRIBE or update options visit:<br> <a href="http://lists.digium.com/mailman/listinfo/asterisk-dev">http://lists.digium.com/mailman/listinfo/asterisk-dev
</a><br></blockquote></div><br>