<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<p>Hi Greg,</p>
<p>The only math library, other than what openHPSDR developers have
written, that we're using today for PowerSDR and follow-ons, if
the FFTW library for FFTs (it's VERY good by the way, FFTs seem
almost free). In WDSP, I've written and included various matrix
inversion algorithms and other things as needed. The issues with
the Intel libraries in the past have been licensing. I think,
over the years, they've softened their stance, somewhat at least,
on that. It deserves a closer look. The benchmark for "Singular
Value Decomposition" caught my eye ... I have something in mind
for that and haven't had time to write the code myself.<span
class="moz-smiley-s1"><span>:-)</span></span></p>
<p>73,</p>
<p>Warren NR0V<br>
</p>
<br>
<div class="moz-cite-prefix">On 5/27/2016 8:23 AM, Gregory W.
Ratcliff wrote:<br>
</div>
<blockquote
cite="mid:1923269153.404686.1464362617770.JavaMail.yahoo@mail.yahoo.com"
type="cite">
<div style="color:#000; background-color:#fff;
font-family:HelveticaNeue-Light, Helvetica Neue Light, Helvetica
Neue, Helvetica, Arial, Lucida Grande,
sans-serif;font-size:16px">
<div id="yui_3_16_0_ym19_1_1464362429373_3916"><span></span></div>
<div id="yui_3_16_0_ym19_1_1464362429373_3916"><span
id="yui_3_16_0_ym19_1_1464362429373_3927">This caught my eye
today as I was working through some big data performance
problems.</span></div>
<div id="yui_3_16_0_ym19_1_1464362429373_3916"><span><br>
</span></div>
<div id="yui_3_16_0_ym19_1_1464362429373_3916" dir="ltr"><span
id="yui_3_16_0_ym19_1_1464362429373_3979">I took the
appliance operator path and installed binaries before I went
digging into source and the API.</span></div>
</div>
</blockquote>
</body>
</html>