sdk/1.1/docs/man/cl_khr_fp64.xml

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook MathML Module V1.1b1//EN"
              "http://www.oasis-open.org/docbook/xml/mathml/1.1CR1/dbmathml.dtd">
<refentry>
    <refentryinfo>
        <keywordset>
            <keyword>cl_khr_fp64</keyword>
        </keywordset>
    </refentryinfo>

    <refmeta>                                    
        <refentrytitle>cl_khr_fp64</refentrytitle>

        <refmiscinfo>
            <copyright>
                <year>2007-2010</year>
                <holder>The Khronos Group Inc. 
 Permission is hereby granted, free of charge, to any person obtaining a
copy of this software and/or associated documentation files (the
"Materials"), to deal in the Materials without restriction, including
without limitation the rights to use, copy, modify, merge, publish,
distribute, sublicense, and/or sell copies of the Materials, and to
permit persons to whom the Materials are furnished to do so, subject to
the condition that this copyright notice and permission notice shall be included
in all copies or substantial portions of the Materials.</holder>
            </copyright>
        </refmiscinfo>
        <manvolnum>3</manvolnum>
    </refmeta>

<!-- ================================ SYNOPSIS -->

    <refnamediv id="cl_khr_fp64">             
        <refname>cl_khr_fp64</refname>

        <refpurpose>
            Support for <type>double</type> floating-point precision.
        </refpurpose>
    </refnamediv>

    <!-- ALTERNATIVE SYNTAX SYNOPSIS (NON-FUNCTION) -->
    <refsect2 id="synopsis">
        <title>
        </title>

        <informaltable frame="none">
            <tgroup cols="1" align="left" colsep="0" rowsep="0">
                <colspec colname="col1" colnum="1" />
                <tbody>
                    <row>
                        <entry>
#pragma OPENCL EXTENSION cl_khr_fp64 : enable
                        </entry>
                    </row>
                </tbody>
            </tgroup>
        </informaltable>
    </refsect2>


<!-- ================================ DESCRIPTION  -->

    <refsect1 id="description"><title>Description</title>
        <para>
Support for double floating-point precision is a requirement for a class of scientific computing algorithms/applications. This class of applications can be enabled by adding support for double precision floating-point as an optional extension.
        </para>
        <para>
OpenCL 1.1 supports double precision floating-point as an optional extension. An application that wants to use <type>double</type> will need to include the <function>#pragma OPENCL EXTENSION cl_khr_fp64 : enable</function> directive before any double precision data type is declared in the kernel code.
        </para>
        <para>
The lists of built-in <citerefentry href="scalarDataTypes"><refentrytitle>scalar data types</refentrytitle></citerefentry> and <citerefentry href="vectorDataTypes"><refentrytitle>vector data types</refentrytitle></citerefentry> are extended to include the following:
            <informaltable frame="all">
                <tgroup cols="2" align="left" colsep="1" rowsep="1">
                    <colspec colname="col1" colnum="1" />
                    <colspec colname="col2" colnum="2" />
                    <colspec colname="col3" colnum="3" />

                    <thead>
                        <row>
                            <entry>Type in OpenCL Language</entry>
                            <entry>Description</entry>
                            <entry>API type for application</entry>
                        </row>
                    </thead>

                    <tbody>
                        <row>
                            <entry>double</entry>
                            <entry>A double precision vector</entry>
                            <entry>cl_double</entry>
                        </row>                  
                        <row>
                            <entry>double2</entry>
                            <entry>2-component double vector</entry>
                            <entry>cl_double2</entry>
                        </row>
                        <row>
                            <entry>double3</entry>
                            <entry>3-component double vector</entry>
                            <entry>cl_double3</entry>
                        </row>
                        <row>
                            <entry>double4</entry>
                            <entry>4-component double vector</entry>
                            <entry>cl_double4</entry>
                        </row>

                        <row>
                            <entry>double8</entry>
                            <entry>8-component double vector</entry>
                            <entry>cl_double8</entry>
                        </row>

                        <row>
                            <entry>double16</entry>
                            <entry>16-component double vector</entry>
                            <entry>cl_double16</entry>
                        </row>

                    </tbody>
                </tgroup>
            </informaltable>
        </para>

        <para>
The double data type must confirm to the IEEE-754 double precision storage format.
        </para>

        <para>
       Conversions from <type>double</type> to <type>half</type> are correctly rounded. Conversions from <type>half</type> to <type>double</type> are lossless.
        </para>
    </refsect1>


<!-- ================================ EXAMPLE  -->
<!-- DO NOT DELETE IN CASE AN EXAMPLE IS ADDED IN THE FUTURE -->
<!--
    <refsect2 id="example1">
        <title>
            Example
        </title>

        <informaltable frame="none">
            <tgroup cols="1" align="left" colsep="0" rowsep="0">
                <colspec colname="col1" colnum="1" />
                <tbody>
                    <row>
                        <entry>
                            Example goes here - it will be set in "code" type with white space preserved.
                        </entry>
                    </row>
                </tbody>
            </tgroup>
        </informaltable>
    </refsect2>
-->


<!-- ================================ SPECIFICATION  -->
<!-- Set the "uri" attribute in the <olink /> element to the "named destination" for the PDF page
-->
    <refsect1 id="specification"><title>Specification</title>
        <para>
            <imageobject>
                <imagedata fileref="pdficon_small1.gif" format="gif" />
            </imageobject>
                                                 
            <olink uri="cl_khr_fp64">OpenCL Specification</olink>
        </para>
    </refsect1>


<!-- ================================ ALSO SEE  -->

    <refsect1 id="seealso"><title>Also see</title>
        <para>
            <citerefentry>
                <refentrytitle>clGetDeviceInfo</refentrytitle>, <citerefentry><refentrytitle>EXTENSION</refentrytitle></citerefentry>
            </citerefentry>
        </para>
    </refsect1>



<!-- ============================== COPYRIGHT -->
<!-- Content included from copyright.inc.xsl -->

    <refsect3 id="Copyright"><title></title>
        <imageobject>
                <imagedata fileref="KhronosLogo.jpg" format="jpg" />
        </imageobject>
        <para />
    </refsect3>

</refentry>