<html>
<head>
<meta content="text/html; charset=utf-8" http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
One possible solution would be to use 2 nvidia cards, say a pair of
gtx750 TIs. At $129 each your<br>
total investment would be much less than the gtx980 ($650).<br>
<br>
I *think* this would make for a relatively easy CUDA sdk install...
but we wont know till someone<br>
actually tries it!<br>
<br>
73,<br>
Steve AD0ES<br>
<br>
<div class="moz-cite-prefix">On 05/12/2016 09:14 AM, Scott Traurig
wrote:<br>
</div>
<blockquote
cite="mid:CAGK9zXorXnYjLVn4tgh37anfOu+5pR-=7ryunSY6wmcS_Uoddg@mail.gmail.com"
type="cite">
<div dir="ltr">Ha ha! Cats and dogs living together ;-) Yes, that
is a recipe for disaster, or at least a lot of skull sweat,
particularly in a linux environment.
<div><br>
</div>
<div>Sadly, I am also using an AMD product for my triple screen
setup. An XFX R7-260X-CNF4 Core Edition Radeon R7 260X 2GB,
plus a Club 3D Displayport mux. I had gone over to AMD back in
the legacy Flex days when the older NVIDIA drivers were
causing poor DPC performance. I've been very happy with the
performance of that setup. But it means that when we are
finally faced with having to support CUDA processing
requirements I'm going to have to go buy a new NVIDIA GPU to
replace the AMD :-(</div>
<div><br>
</div>
<div>73!</div>
<div><br>
</div>
<div>Scott/w-u-2-o</div>
<div><br>
</div>
<div><br>
<div class="gmail_extra"><br>
<div class="gmail_quote">On Thu, May 12, 2016 at 11:01 AM,
ad0es <span dir="ltr"><<a moz-do-not-send="true"
href="mailto:ad0es@ad0es.net" target="_blank">ad0es@ad0es.net</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="margin:0px 0px 0px
0.8ex;border-left-width:1px;border-left-style:solid;border-left-color:rgb(204,204,204);padding-left:1ex">*****
High Performance Software Defined Radio Discussion List
*****<br>
<br>
For my experiments I chose to avoid the sharing issue.
I use an AMD eyefinity 6 card for video, and a gtx980
for CUDA.<br>
This introduces problems of its own:<br>
<br>
<a moz-do-not-send="true"
href="http://www.ad0es.net/hpsdr/ngSDR/cudainstall/index.html"
rel="noreferrer" target="_blank">http://www.ad0es.net/hpsdr/ngSDR/cudainstall/index.html</a><br>
<br>
On 05/12/2016 08:40 AM, Scott Traurig wrote:<br>
<blockquote class="gmail_quote" style="margin:0px 0px
0px
0.8ex;border-left-width:1px;border-left-style:solid;border-left-color:rgb(204,204,204);padding-left:1ex">*****
High Performance Software Defined Radio Discussion
List *****<br>
My understanding is that as long as a reasonably
serious NVIDIA graphics card is installed for video,
sufficient CUDA processing elements can be borrowed
from it without affecting video functionality and
without requiring a second, dedicated GPU.<br>
<br>
<br>
</blockquote>
<br>
_______________________________________________<br>
HPSDR Discussion List<br>
To post msg: <a moz-do-not-send="true"
href="mailto:hpsdr@openhpsdr.org" target="_blank">hpsdr@openhpsdr.org</a><br>
Subscription help: <a moz-do-not-send="true"
href="http://lists.openhpsdr.org/listinfo.cgi/hpsdr-openhpsdr.org"
rel="noreferrer" target="_blank">http://lists.openhpsdr.org/listinfo.cgi/hpsdr-openhpsdr.org</a><br>
HPSDR web page: <a moz-do-not-send="true"
href="http://openhpsdr.org" rel="noreferrer"
target="_blank">http://openhpsdr.org</a><br>
Archives: <a moz-do-not-send="true"
href="http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/"
rel="noreferrer" target="_blank">http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/</a><br>
</blockquote>
</div>
<br>
</div>
</div>
</div>
</blockquote>
<br>
</body>
</html>