<!DOCTYPE html
PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" lang="en-us" xml:lang="en-us">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8"></meta>
<meta http-equiv="X-UA-Compatible" content="IE=edge"></meta>
<meta name="copyright" content="(C) Copyright 2005"></meta>
<meta name="DC.rights.owner" content="(C) Copyright 2005"></meta>
<meta name="DC.Type" content="concept"></meta>
<meta name="DC.Title" content="NVIDIA CUDA Getting Started Guide for Mac OS X"></meta>
<meta name="abstract" content="The installation instructions for the CUDA Toolkit on Mac OS X."></meta>
<meta name="description" content="The installation instructions for the CUDA Toolkit on Mac OS X."></meta>
<meta name="DC.Coverage" content="Getting Started Guides"></meta>
<meta name="DC.subject" content="CUDA Mac, CUDA Mac requirements, CUDA Mac supported versions, CUDA Mac gcc, CUDA Mac driver, CUDA Mac GPU, CUDA Mac download, CUDA Mac install verification, CUDA Mac examples, CUDA Mac binaries"></meta>
<meta name="keywords" content="CUDA Mac, CUDA Mac requirements, CUDA Mac supported versions, CUDA Mac gcc, CUDA Mac driver, CUDA Mac GPU, CUDA Mac download, CUDA Mac install verification, CUDA Mac examples, CUDA Mac binaries"></meta>
<meta name="DC.Format" content="XHTML"></meta>
<meta name="DC.Identifier" content="abstract"></meta>
<link rel="stylesheet" type="text/css" href="../common/formatting/commonltr.css"></link>
<link rel="stylesheet" type="text/css" href="../common/formatting/site.css"></link>
<title>Getting Started Mac OS X :: CUDA Toolkit Documentation</title>
<!--[if lt IE 9]>
<script src="../common/formatting/html5shiv-printshiv.min.js"></script>
<![endif]-->
<script type="text/javascript" charset="utf-8" src="../common/formatting/jquery.min.js"></script>
<script type="text/javascript" charset="utf-8" src="../common/formatting/jquery.ba-hashchange.min.js"></script>
<link rel="canonical" href="http://docs.nvidia.com/cuda/cuda-getting-started-guide-for-mac-os-x/index.html"></link>
<link rel="stylesheet" type="text/css" href="../common/formatting/qwcode.highlight.css"></link>
</head>
<body>
<article id="contents">
<div id="release-info">Getting Started Mac OS X
(<a href="../../pdf/CUDA_Getting_Started_Mac.pdf">PDF</a>)
-
CUDA Toolkit v5.5
(<a href="https://developer.nvidia.com/cuda-toolkit-archive">older</a>)
-
Last updated
July 19, 2013
-
<a href="mailto:cudatools@nvidia.com?subject=CUDA Tools Documentation Feedback: cuda-getting-started-guide-for-mac-os-x">Send Feedback</a></div>
<div class="topic nested0" id="abstract"><a name="abstract" shape="rect">
<!-- --></a><h2 class="title topictitle1"><a href="#abstract" name="abstract" shape="rect">NVIDIA CUDA Getting Started Guide for Mac OS X</a></h2>
<div class="body conbody"></div>
</div>
<div class="topic concept nested0" id="introduction"><a name="introduction" shape="rect">
<!-- --></a><h2 class="title topictitle1"><a href="#introduction" name="introduction" shape="rect">1. Introduction</a></h2>
<div class="body conbody">
<p class="p">CUDA™ is a parallel computing platform and programming model invented by NVIDIA. It enables dramatic increases in computing
performance by harnessing the power of the graphics processing unit (GPU).
</p>
<div class="p">CUDA was developed with several design goals in mind:
<ul class="ul">
<li class="li">Provide a small set of extensions to standard programming languages, like C, that enable a straightforward implementation
of parallel algorithms. With CUDA C/C++, programmers can focus on the task of parallelization of the algorithms rather than
spending time on their implementation.
</li>
<li class="li">Support heterogeneous computation where applications use both the CPU and GPU. Serial portions of applications are run on
the CPU, and parallel portions are offloaded to the GPU. As such, CUDA can be incrementally applied to existing applications.
The CPU and GPU are treated as separate devices that have their own memory spaces. This configuration also allows simultaneous
computation on the CPU and GPU without contention for memory resources.
</li>
</ul>
CUDA-capable GPUs have hundreds of cores that can collectively run thousands of computing threads. These cores have shared
resources including a register file and a shared memory. The on-chip shared memory allows parallel tasks running on these
cores to share data without sending it over the system memory bus.
</div>
<p class="p">This guide will show you how to install and check the correct operation of the CUDA development tools.</p>
</div>
<div class="topic concept nested1" id="system-requirements"><a name="system-requirements" shape="rect">
<!-- --></a><h3 class="title topictitle2"><a href="#system-requirements" name="system-requirements" shape="rect">1.1. System Requirements</a></h3>
<div class="body conbody">
<div class="p">To use CUDA on your system, you will need the following installed:
<ul class="ul">
<li class="li">CUDA-capable GPU</li>
<li class="li">Mac OSX v. 10.7.5 or later</li>
<li class="li">The <dfn class="term">gcc</dfn> or <dfn class="term">Clang</dfn> compiler and toolchain installed using Xcode
</li>
<li class="li">NVIDIA CUDA Toolkit (available at <a class="xref" href="http://developer.nvidia.com/cuda-downloads" target="_blank" shape="rect">http://developer.nvidia.com/cuda-downloads</a>)
</li>
</ul>
</div>
</div>
</div>
<div class="topic concept nested1" id="about-this-document"><a name="about-this-document" shape="rect">
<!-- --></a><h3 class="title topictitle2"><a href="#about-this-document" name="about-this-document" shape="rect">1.2. About This Document</a></h3>
<div class="body conbody">
<p class="p">This document is intended for readers familiar with the Mac OS X environment and the compilation of C programs from the command
line. You do not need previous experience with CUDA or experience with parallel computation.
</p>
</div>
</div>
</div>
<div class="topic concept nested0" id="installing-cuda-development-tools"><a name="installing-cuda-development-tools" shape="rect">
<!-- --></a><h2 class="title topictitle1"><a href="#installing-cuda-development-tools" name="installing-cuda-development-tools" shape="rect">2. Installing CUDA Development Tools</a></h2>
<div class="body conbody">
<div class="p">The setup of CUDA development tools on a system running
Mac OS X consists of a few simple steps:
<ul class="ul">
<li class="li">Verify the system has a CUDA-capable GPU.</li>
<li class="li">Verify the system is running a supported version of Mac OS X.</li>
<li class="li">Verify the system has gcc or Clang installed via Xcode.</li>
<li class="li">Download the NVIDIA CUDA Toolkit.</li>
<li class="li">Install the NVIDIA CUDA Toolkit.</li>
<li class="li">Test that the installed software runs correctly and communicates with the hardware.</li>
</ul>
</div>
</div>
<div class="topic concept nested1" id="verify-you-have-cuda-enabled-system"><a name="verify-you-have-cuda-enabled-system" shape="rect">
<!-- --></a><h3 class="title topictitle2"><a href="#verify-you-have-cuda-enabled-system" name="verify-you-have-cuda-enabled-system" shape="rect">2.1. Verify You Have a CUDA-Capable GPU</a></h3>
<div class="body conbody">
<p class="p">To verify that your system is CUDA-capable, under the <span class="keyword wintitle">Apple</span> menu select <span class="ph uicontrol">About This Mac</span>, click the <span class="ph uicontrol">More Info …</span> button, and then select <span class="ph uicontrol">Graphics/Displays</span> under the <span class="keyword wintitle">Hardware</span> list. There you will find the vendor name and model of your graphics card.
If it is an NVIDIA card that is listed in <a class="xref" href="http://developer.nvidia.com/cuda-gpus" target="_blank" shape="rect">http://developer.nvidia.com/cuda-gpus</a>, your GPU is CUDA-capable.
</p>
<p class="p">The Release Notes for the CUDA Toolkit also contain a list of supported products.</p>
</div>
</div>
<div class="topic concept nested1" id="verify-correct-version-of-mac-os-x"><a name="verify-correct-version-of-mac-os-x" shape="rect">
<!-- --></a><h3 class="title topictitle2"><a href="#verify-correct-version-of-mac-os-x" name="verify-correct-version-of-mac-os-x" shape="rect">2.2. Verify the Correct Version of Mac OS X</a></h3>
<div class="body conbody">
<p class="p">The CUDA Development Tools require an Intel-based Mac running Mac OSX v. 10.7.5 or later. To check which version you have,
go to the <span class="keyword wintitle">Apple</span> menu on the desktop and select <span class="ph uicontrol">About This Mac</span>. You should see a dialog box similar to <a class="xref" href="index.html#verify-correct-version-of-mac-os-x__macosx" shape="rect">Figure 1</a>.
</p>
<div class="fig fignone" id="verify-correct-version-of-mac-os-x__macosx"><a name="verify-correct-version-of-mac-os-x__macosx" shape="rect">
<!-- --></a><span class="figcap">Figure 1. About This Mac Dialog Box</span><br clear="none"></br><div class="imageleft"><img class="image imageleft" src="graphics/about-this-mac-dialog-box.png" alt="About This Mac Dialog Box."></img></div><br clear="none"></br></div>
</div>
<div class="topic concept nested2" id="verify-gcc-installed"><a name="verify-gcc-installed" shape="rect">
<!-- --></a><h3 class="title topictitle2"><a href="#verify-gcc-installed" name="verify-gcc-installed" shape="rect">Verify the System Has gcc Installed</a></h3>
<div class="body conbody">
<p class="p">The <samp class="ph codeph">gcc</samp> compiler and toolchain are installed using the installation of Xcode. The Xcode development environment is found on the <cite class="cite">Xcode Developer Tools DVD</cite> that ships with new Mac systems and with Leopard, if you buy the operating-system upgrade. When installing Xcode, the package
that contains gcc and the necessary tools is called <cite class="cite">Developer Tools Essentials</cite>. You can verify that gcc is installed entering the command <samp class="ph codeph">/usr/bin/gcc –-help</samp> from a <span class="keyword wintitle">Terminal</span> window.
</p>
</div>
</div>
</div>
<div class="topic concept nested1" id="download-cuda-software"><a name="download-cuda-software" shape="rect">
<!-- --></a><h3 class="title topictitle2"><a href="#download-cuda-software" name="download-cuda-software" shape="rect">2.3. Download the CUDA Software</a></h3>
<div class="body conbody">
<p class="p">Once you have verified that you have a supported NVIDIA GPU, a supported version the MAC OS, and gcc, you need to download
the NVIDIA CUDA Toolkit.
</p>
<p class="p">The NVIDIA CUDA Toolkit is available at no cost from the main CUDA download site at <a class="xref" href="http://www.nvidia.com/content/cuda/cuda-downloads.html" target="_blank" shape="rect">http://www.nvidia.com/content/cuda/cuda-downloads.html</a>. It contains the driver and tools needed to create, build and run a CUDA application as well as libraries, header files,
CUDA samples source code, and other resources.
</p>
</div>
</div>
<div class="topic task nested1" id="install-cuda-driver-and-software"><a name="install-cuda-driver-and-software" shape="rect">
<!-- --></a><h3 class="title topictitle2"><a href="#install-cuda-driver-and-software" name="install-cuda-driver-and-software" shape="rect">2.4. Install the CUDA Driver and Software</a></h3>
<div class="body taskbody">
<div class="section context">
<p class="p">Use the following procedure to successfully install the CUDA driver and software. For information not listed here, see the
documentation under <samp class="ph codeph">/Developer/NVIDIA/CUDA-5.5/doc</samp> in the download location.
</p>
<p class="p">Before installing the CUDA Toolkit, you should read the <cite class="cite">Release Notes</cite>, as they provide important details on installation and software functionality.
</p>
<p class="p">Then, follow these few steps for a successful installation.</p>
</div>
<ol class="ol steps">
<li class="li step">
<div class="note note"><span class="notetitle">Note:</span> The driver and toolkit must be installed for CUDA to function. If you have not installed a stand-alone driver, install the
driver from the NVIDIA CUDA Toolkit.
</div><span class="ph cmd">Install the CUDA Toolkit.</span>
Install the CUDA toolkit by executing the installer and following the on-screen prompts. You will be able to choose which
packages you wish to install. The packages are:
<ul class="ul">
<li class="li"><strong class="ph b">CUDA Driver</strong><p class="p">This will install <samp class="ph codeph">/Library/Frameworks/CUDA.framework</samp> and the UNIX-compatibility stub <samp class="ph codeph">/usr/local/cuda/lib/libcuda.dylib</samp> that refers to it.
</p>
</li>
<li class="li"><strong class="ph b">CUDA Toolkit</strong><p class="p">The CUDA Toolkit supplements the CUDA Driver with compilers and additional libraries and header files that are installed into
<samp class="ph codeph">/Developer/NVIDIA/CUDA-5.5</samp> by default. Symlinks are created in <samp class="ph codeph">/usr/local/cuda/</samp> pointing to their respective files in <samp class="ph codeph">/Developer/NVIDIA/CUDA-5.5/</samp> .
</p>
<p class="p">Previous installations of the toolkit will be moved to <samp class="ph codeph">/Developer/NVIDIA/CUDA-#.#</samp> to better support side-by-side installations.
</p>
</li>
<li class="li"><strong class="ph b">CUDA Samples</strong><p class="p">The CUDA Samples are installed to <samp class="ph codeph">/Developer/NVIDIA/CUDA-5.5/samples</samp>.
</p>
<p class="p">Previous installations of the samples will be moved to <samp class="ph codeph">/Developer/NVIDIA/CUDA-#.#/samples</samp> to better support side-by-side installations.
</p>
</li>
</ul>
</li>
<li class="li step"><span class="ph cmd">Define the environment variables.</span><ul class="ul">
<li class="li">The <samp class="ph codeph">PATH</samp> variable needs to include <samp class="ph codeph">/Developer/NVIDIA/CUDA-5.5/bin</samp></li>
<li class="li"><samp class="ph codeph">DYLD_LIBRARY_PATH</samp> needs to contain <samp class="ph codeph">/Developer/NVIDIA/CUDA-5.5/lib</samp></li>
</ul>
<p class="p">To change the environment variables for 32-bit operating systems:</p><pre class="pre screen" xml:space="preserve">export PATH=/Developer/NVIDIA/CUDA-5.5/bin:$PATH
export DYLD_LIBRARY_PATH=/Developer/NVIDIA/CUDA-5.5/lib:$DYLD_LIBRARY_PATH</pre><p class="p">To change the environment variables for 64-bit operating systems:</p><pre class="pre screen" xml:space="preserve">export PATH=/Developer/NVIDIA/CUDA-5.5/bin:$PATH
export DYLD_LIBRARY_PATH=/Developer/NVIDIA/CUDA-5.5/lib:$DYLD_LIBRARY_PATH</pre></li>
<li class="li step"><span class="ph cmd">(Optional) Install a writable copy of the samples.</span>
In order to modify, compile, and run the samples, the samples must be installed with write permissions. A convenience installation
script is provided:
<pre class="pre screen" xml:space="preserve">cuda-install-samples-5.5.sh <dir></pre>
This script is installed with the cuda-samples-5-5 package. The cuda-samples-5-5 package installs only a read-only copy in
/Developer/NVIDIA/CUDA-5.5/samples.
</li>
</ol>
</div>
</div>
<div class="topic concept nested1" id="verify-installation"><a name="verify-installation" shape="rect">
<!-- --></a><h3 class="title topictitle2"><a href="#verify-installation" name="verify-installation" shape="rect">2.5. Verify the Installation</a></h3>
<div class="body conbody">
<p class="p">Before continuing, it is important to verify that the CUDA toolkit can find and communicate correctly with the CUDA-capable
hardware. To do this, you need to compile and run some of the included sample programs.
</p>
<div class="note note"><span class="notetitle">Note:</span> Ensure the PATH and DYLD_LIBRARY_PATH variables are <a class="xref" href="index.html#install-cuda-driver-and-software" shape="rect">set correctly</a>.
</div>
</div>
<div class="topic concept nested2" id="verify-driver-installation"><a name="verify-driver-installation" shape="rect">
<!-- --></a><h3 class="title topictitle2"><a href="#verify-driver-installation" name="verify-driver-installation" shape="rect">2.5.1. Verify the Driver Installation</a></h3>
<div class="body conbody">
<div class="p">If the CUDA Driver is installed correctly, the CUDA kernel extension (<samp class="ph codeph">/System/Library/Extensions/CUDA.kext</samp>) should be loaded automatically at boot time. To verify that it is loaded, use the command
<pre class="pre screen" xml:space="preserve">kextstat | grep -i cuda</pre></div>
</div>
</div>
<div class="topic concept nested2" id="compiling-examples"><a name="compiling-examples" shape="rect">
<!-- --></a><h3 class="title topictitle2"><a href="#compiling-examples" name="compiling-examples" shape="rect">2.5.2. Compiling the Examples</a></h3>
<div class="body conbody">
<p class="p">The version of the CUDA Toolkit can be checked by running <samp class="ph codeph">nvcc -V</samp> in a terminal window. The <samp class="ph codeph">nvcc</samp> command runs the compiler driver that compiles CUDA programs. It calls the <samp class="ph codeph">gcc</samp> compiler for C code and the NVIDIA PTX compiler for the CUDA code.
</p>
<p class="p">The NVIDIA CUDA Toolkit includes sample programs in source form. You should compile them by changing to the directory where
you installed the writable samples,<samp class="ph codeph"><dir></samp>, and typing <samp class="ph codeph">make</samp>. The resulting binaries will be placed under <samp class="ph codeph"><dir>/bin/x86_64/darwin/release</samp>.
</p>
</div>
</div>
<div class="topic concept nested2" id="running-binaries"><a name="running-binaries" shape="rect">
<!-- --></a><h3 class="title topictitle2"><a href="#running-binaries" name="running-binaries" shape="rect">2.5.3. Running Binaries</a></h3>
<div class="body conbody">
<p class="p">After compilation, go to <samp class="ph codeph"><dir>/bin/x86_64/darwin/release</samp> and run <samp class="ph codeph">deviceQuery</samp>. If the CUDA software is installed and configured correctly, the output for <samp class="ph codeph">deviceQuery</samp> should look similar to that shown in <a class="xref" href="index.html#running-binaries__valid-results-from-sample-cuda-devicequery-program" shape="rect">Figure 2</a>.
</p>
<div class="fig fignone" id="running-binaries__valid-results-from-sample-cuda-devicequery-program"><a name="running-binaries__valid-results-from-sample-cuda-devicequery-program" shape="rect">
<!-- --></a><span class="figcap">Figure 2. Valid Results from deviceQuery CUDA Sample</span><br clear="none"></br><div class="imagecenter"><img class="image imagecenter" src="graphics/valid-results-from-sample-cuda-devicequery-program.png" width="550" alt="Valid Results from deviceQuery CUDA Sample."></img></div><br clear="none"></br></div>
<p class="p">Note that the parameters for your CUDA device will vary. The key lines are the first and second ones that confirm a device
was found and what model it is. Also, the next-to-last line, as indicated, should show that the test passed.
</p>
<p class="p">Running the <samp class="ph codeph">bandwidthTest</samp> program ensures that the system and the CUDA-capable device are able to communicate correctly. Its output is shown in <a class="xref" href="index.html#running-binaries__valid-results-from-sample-cuda-bandwidthtest-program" shape="rect">Figure 3</a>.
</p>
<div class="fig fignone" id="running-binaries__valid-results-from-sample-cuda-bandwidthtest-program"><a name="running-binaries__valid-results-from-sample-cuda-bandwidthtest-program" shape="rect">
<!-- --></a><span class="figcap">Figure 3. Valid Results from bandwidthTest CUDA Sample</span><br clear="none"></br><div class="imagecenter"><img class="image imagecenter" src="graphics/valid-results-from-sample-cuda-bandwidthtest-program.png" width="550" alt="Valid Results from bandwidthTest CUDA Sample."></img></div><br clear="none"></br></div>
<p class="p">Note that the measurements for your CUDA-capable device description will vary from system to system. The important point is
that you obtain measurements, and that the second-to-last line (in <a class="xref" href="index.html#running-binaries__valid-results-from-sample-cuda-bandwidthtest-program" shape="rect">Figure 3</a>) confirms that all necessary tests passed.
</p>
<p class="p">Should the tests not pass, make sure you have a CUDA-capable NVIDIA GPU on your system and make sure it is properly installed.
</p>
<p class="p">If you run into difficulties with the link step (such as libraries not being found), consult the <cite class="cite">Release Notes</cite> found in the <samp class="ph codeph">doc</samp> folder in the CUDA Samples directory.
</p>
<p class="p">To see a graphical representation of what CUDA can do, run the particles executable.</p>
</div>
</div>
</div>
</div>
<div class="topic concept nested0" id="additional-considerations"><a name="additional-considerations" shape="rect">
<!-- --></a><h2 class="title topictitle1"><a href="#additional-considerations" name="additional-considerations" shape="rect">3. Additional Considerations</a></h2>
<div class="body conbody">
<p class="p">Now that you have CUDA-capable hardware and the NVIDIA CUDA Toolkit installed, you can examine and enjoy the numerous included
programs. To begin using CUDA to accelerate the performance of your own applications, consult the <cite class="cite">CUDA C Programming Guide</cite>, located in <samp class="ph codeph">/Developer/NVIDIA/CUDA-5.5/doc</samp>.
</p>
<p class="p">A number of helpful development tools are included in the CUDA Toolkit to assist you as you develop your CUDA programs, such
as NVIDIA<sup>®</sup> Nsight™ Eclipse Edition, NVIDIA Visual Profiler, cuda-gdb, and cuda-memcheck.
</p>
<p class="p">For technical support on programming questions, consult and participate in the developer forums at <a class="xref" href="http://developer.nvidia.com/cuda/" target="_blank" shape="rect">http://developer.nvidia.com/cuda/</a>.
</p>
</div>
</div>
<div class="topic concept nested0" id="notices-header"><a name="notices-header" shape="rect">
<!-- --></a><h2 class="title topictitle1"><a href="#notices-header" name="notices-header" shape="rect">Notices</a></h2>
<div class="topic reference nested1" id="notice"><a name="notice" shape="rect">
<!-- --></a><h3 class="title topictitle2"><a href="#notice" name="notice" shape="rect"></a></h3>
<div class="body refbody">
<div class="section">
<h3 class="title sectiontitle">Notice</h3>
<p class="p">ALL NVIDIA DESIGN SPECIFICATIONS, REFERENCE BOARDS, FILES, DRAWINGS, DIAGNOSTICS, LISTS, AND OTHER DOCUMENTS (TOGETHER AND
SEPARATELY, "MATERIALS") ARE BEING PROVIDED "AS IS." NVIDIA MAKES NO WARRANTIES, EXPRESSED, IMPLIED, STATUTORY, OR OTHERWISE
WITH RESPECT TO THE MATERIALS, AND EXPRESSLY DISCLAIMS ALL IMPLIED WARRANTIES OF NONINFRINGEMENT, MERCHANTABILITY, AND FITNESS
FOR A PARTICULAR PURPOSE.
</p>
<p class="p">Information furnished is believed to be accurate and reliable. However, NVIDIA Corporation assumes no responsibility for the
consequences of use of such information or for any infringement of patents or other rights of third parties that may result
from its use. No license is granted by implication of otherwise under any patent rights of NVIDIA Corporation. Specifications
mentioned in this publication are subject to change without notice. This publication supersedes and replaces all other information
previously supplied. NVIDIA Corporation products are not authorized as critical components in life support devices or systems
without express written approval of NVIDIA Corporation.
</p>
</div>
</div>
</div>
<div class="topic reference nested1" id="trademarks"><a name="trademarks" shape="rect">
<!-- --></a><h3 class="title topictitle2"><a href="#trademarks" name="trademarks" shape="rect"></a></h3>
<div class="body refbody">
<div class="section">
<h3 class="title sectiontitle">Trademarks</h3>
<p class="p">NVIDIA and the NVIDIA logo are trademarks or registered trademarks of NVIDIA Corporation
in the U.S. and other countries. Other company and product names may be trademarks of
the respective companies with which they are associated.
</p>
</div>
</div>
</div>
<div class="topic reference nested1" id="copyright-past-to-present"><a name="copyright-past-to-present" shape="rect">
<!-- --></a><h3 class="title topictitle2"><a href="#copyright-past-to-present" name="copyright-past-to-present" shape="rect"></a></h3>
<div class="body refbody">
<div class="section">
<h3 class="title sectiontitle">Copyright</h3>
<p class="p">© <span class="ph">2009</span>-<span class="ph">2013</span> NVIDIA
Corporation. All rights reserved.
</p>
</div>
</div>
</div>
</div>
<hr id="contents-end"></hr>
<div id="release-info">Getting Started Mac OS X
(<a href="../../pdf/CUDA_Getting_Started_Mac.pdf">PDF</a>)
-
CUDA Toolkit v5.5
(<a href="https://developer.nvidia.com/cuda-toolkit-archive">older</a>)
-
Last updated
July 19, 2013
-
<a href="mailto:cudatools@nvidia.com?subject=CUDA Tools Documentation Feedback: cuda-getting-started-guide-for-mac-os-x">Send Feedback</a></div>
</article>
<header id="header"><span id="company">NVIDIA</span><span id="site-title">CUDA Toolkit Documentation</span><form id="search" method="get" action="search">
<input type="text" name="search-text"></input><fieldset id="search-location">
<legend>Search In:</legend>
<label><input type="radio" name="search-type" value="site"></input>Entire Site</label>
<label><input type="radio" name="search-type" value="document"></input>Just This Document</label></fieldset>
<button type="reset">clear search</button>
<button id="submit" type="submit">search</button></form>
</header>
<nav id="site-nav">
<div class="category closed"><span class="twiddle">▷</span><a href="../index.html" title="The root of the site.">CUDA Toolkit</a></div>
<ul class="closed">
<li><a href="../cuda-toolkit-release-notes/index.html" title="The Release Notes for the CUDA Toolkit from v4.0 to today.">Release Notes</a></li>
<li><a href="../eula/index.html" title="The End User License Agreements for the NVIDIA CUDA Toolkit, the NVIDIA CUDA Samples, the NVIDIA Display Driver, and NVIDIA NSight (Visual Studio Edition).">EULA</a></li>
<li><a href="../cuda-getting-started-guide-for-linux/index.html" title="This guide discusses how to install and check for correct operation of the CUDA Development Tools on GNU/Linux systems.">Getting Started Linux</a></li>
<li><a href="../cuda-getting-started-guide-for-mac-os-x/index.html" title="This guide discusses how to install and check for correct operation of the CUDA Development Tools on Mac OS X systems.">Getting Started Mac OS X</a></li>
<li><a href="../cuda-getting-started-guide-for-microsoft-windows/index.html" title="This guide discusses how to install and check for correct operation of the CUDA Development Tools on Microsoft Windows systems.">Getting Started Windows</a></li>
<li><a href="../cuda-c-programming-guide/index.html" title="This guide provides a detailed discussion of the CUDA programming model and programming interface. It then describes the hardware implementation, and provides guidance on how to achieve maximum performance. The Appendixes include a list of all CUDA-enabled devices, detailed description of all extensions to the C language, listings of supported mathematical functions, C++ features supported in host and device code, details on texture fetching, technical specifications of various devices, and concludes by introducing the low-level driver API.">Programming Guide</a></li>
<li><a href="../cuda-c-best-practices-guide/index.html" title="This guide presents established parallelization and optimization techniques and explains coding metaphors and idioms that can greatly simplify programming for CUDA-capable GPU architectures. The intent is to provide guidelines for obtaining the best performance from NVIDIA GPUs using the CUDA Toolkit.">Best Practices Guide</a></li>
<li><a href="../kepler-compatibility-guide/index.html" title="This application note is intended to help developers ensure that their NVIDIA CUDA applications will run effectively on GPUs based on the NVIDIA Kepler Architecture. This document provides guidance to ensure that your software applications are compatible with Kepler.">Kepler Compatibility Guide</a></li>
<li><a href="../kepler-tuning-guide/index.html" title="Kepler is NVIDIA's next-generation architecture for CUDA compute applications. Applications that follow the best practices for the Fermi architecture should typically see speedups on the Kepler architecture without any code changes. This guide summarizes the ways that an application can be fine-tuned to gain additional speedups by leveraging Kepler architectural features.">Kepler Tuning Guide</a></li>
<li><a href="../parallel-thread-execution/index.html" title="This guide provides detailed instructions on the use of PTX, a low-level parallel thread execution virtual machine and instruction set architecture (ISA). PTX exposes the GPU as a data-parallel computing device.">PTX ISA</a></li>
<li><a href="../optimus-developer-guide/index.html" title="This document explains how CUDA APIs can be used to query for GPU capabilities in NVIDIA Optimus systems.">Developer Guide for Optimus</a></li>
<li><a href="../video-decoder/index.html" title="This document provides the video decoder API specification and the format conversion and display using DirectX or OpenGL following decode.">Video Decoder</a></li>
<li><a href="../video-encoder/index.html" title="This document provides the CUDA video encoder specifications, including the C-library API functions and encoder query parameters.">Video Encoder</a></li>
<li><a href="../inline-ptx-assembly/index.html" title="This document shows how to inline PTX (parallel thread execution) assembly language statements into CUDA code. It describes available assembler statement parameters and constraints, and the document also provides a list of some pitfalls that you may encounter.">Inline PTX Assembly</a></li>
<li><a href="../cuda-runtime-api/index.html" title="The CUDA runtime API.">CUDA Runtime API</a></li>
<li><a href="../cuda-driver-api/index.html" title="The CUDA driver API.">CUDA Driver API</a></li>
<li><a href="../cuda-math-api/index.html" title="The CUDA math API.">CUDA Math API</a></li>
<li><a href="../cublas/index.html" title="The CUBLAS library is an implementation of BLAS (Basic Linear Algebra Subprograms) on top of the NVIDIA CUDA runtime. It allows the user to access the computational resources of NVIDIA Graphical Processing Unit (GPU), but does not auto-parallelize across multiple GPUs.">CUBLAS</a></li>
<li><a href="../cufft/index.html" title="The CUFFT library user guide.">CUFFT</a></li>
<li><a href="../curand/index.html" title="The CURAND library user guide.">CURAND</a></li>
<li><a href="../cusparse/index.html" title="The CUSPARSE library user guide.">CUSPARSE</a></li>
<li><a href="../npp/index.html" title="NVIDIA NPP is a library of functions for performing CUDA accelerated processing. The initial set of functionality in the library focuses on imaging and video processing and is widely applicable for developers in these areas. NPP will evolve over time to encompass more of the compute heavy tasks in a variety of problem domains. The NPP library is written to maximize flexibility, while maintaining high performance.">NPP</a></li>
<li><a href="../thrust/index.html" title="The Thrust getting started guide.">Thrust</a></li>
<li><a href="../cuda-samples/index.html" title="This document contains a complete listing of the code samples that are included with the NVIDIA CUDA Toolkit. It describes each code sample, lists the minimum GPU specification, and provides links to the source code and white papers if available.">CUDA Samples</a></li>
<li><a href="../cuda-compiler-driver-nvcc/index.html" title="This document is a reference guide on the use of the CUDA compiler driver nvcc. Instead of being a specific CUDA compilation driver, nvcc mimics the behavior of the GNU compiler gcc, accepting a range of conventional compiler options, such as for defining macros and include/library paths, and for steering the compilation process.">NVCC</a></li>
<li><a href="../cuda-gdb/index.html" title="The NVIDIA tool for debugging CUDA applications running on Linux and Mac, providing developers with a mechanism for debugging CUDA applications running on actual hardware. CUDA-GDB is an extension to the x86-64 port of GDB, the GNU Project debugger.">CUDA-GDB</a></li>
<li><a href="../cuda-memcheck/index.html" title="CUDA-MEMCHECK is a suite of run time tools capable of precisely detecting out of bounds and misaligned memory access errors, checking device allocation leaks, reporting hardware errors and identifying shared memory data access hazards.">CUDA-MEMCHECK</a></li>
<li><a href="../nsight-eclipse-edition-getting-started-guide/index.html" title="Nsight Eclipse Edition getting started guide">Nsight Eclipse Edition</a></li>
<li><a href="../profiler-users-guide/index.html" title="This is the guide to the Profiler.">Profiler</a></li>
<li><a href="../cuda-binary-utilities/index.html" title="The application notes for cuobjdump and nvdisasm.">CUDA Binary Utilities</a></li>
<li><a href="../floating-point/index.html" title="A number of issues related to floating point accuracy and compliance are a frequent source of confusion on both CPUs and GPUs. The purpose of this white paper is to discuss the most common issues related to NVIDIA GPUs and to supplement the documentation in the CUDA C Programming Guide.">Floating Point and IEEE 754</a></li>
<li><a href="../incomplete-lu-cholesky/index.html" title="In this white paper we show how to use the CUSPARSE and CUBLAS libraries to achieve a 2x speedup over CPU in the incomplete-LU and Cholesky preconditioned iterative methods. We focus on the Bi-Conjugate Gradient Stabilized and Conjugate Gradient iterative methods, that can be used to solve large sparse nonsymmetric and symmetric positive definite linear systems, respectively. Also, we comment on the parallel sparse triangular solve, which is an essential building block in these algorithms.">Incomplete-LU and Cholesky Preconditioned Iterative Methods</a></li>
<li><a href="../libnvvm-api/index.html" title="The libNVVM API.">libNVVM API</a></li>
<li><a href="../libdevice-users-guide/index.html" title="The libdevice library is an LLVM bitcode library that implements common functions for GPU kernels.">libdevice User's Guide</a></li>
<li><a href="../nvvm-ir-spec/index.html" title="NVVM IR is a compiler IR (internal representation) based on the LLVM IR. The NVVM IR is designed to represent GPU compute kernels (for example, CUDA kernels). High-level language front-ends, like the CUDA C compiler front-end, can generate NVVM IR.">NVVM IR</a></li>
<li><a href="../cupti/index.html" title="The CUPTI API.">CUPTI</a></li>
<li><a href="../debugger-api/index.html" title="The CUDA debugger API.">Debugger API</a></li>
<li><a href="../gpudirect-rdma/index.html" title="A tool for Kepler-class GPUs and CUDA 5.0 enabling a direct path for communication between the GPU and a peer device on the PCI Express bus when the devices share the same upstream root complex using standard features of PCI Express. This document introduces the technology and describes the steps necessary to enable a RDMA for GPUDirect connection to NVIDIA GPUs within the Linux device driver model.">RDMA for GPUDirect</a></li>
</ul>
<div class="category"><span class="twiddle">▼</span><a href="index.html" title="Getting Started Mac OS X">Getting Started Mac OS X</a></div>
<ul>
<li><a href="#introduction">1. Introduction</a><ul>
<li><a href="#system-requirements">1.1. System Requirements</a></li>
<li><a href="#about-this-document">1.2. About This Document</a></li>
</ul>
</li>
<li><a href="#installing-cuda-development-tools">2. Installing CUDA Development Tools</a><ul>
<li><a href="#verify-you-have-cuda-enabled-system">2.1. Verify You Have a CUDA-Capable GPU</a></li>
<li><a href="#verify-correct-version-of-mac-os-x">2.2. Verify the Correct Version of Mac OS X</a><ul>
<li><a href="#verify-gcc-installed">2.2.1. Verify the System Has gcc Installed</a></li>
</ul>
</li>
<li><a href="#download-cuda-software">2.3. Download the CUDA Software</a></li>
<li><a href="#install-cuda-driver-and-software">2.4. Install the CUDA Driver and Software</a></li>
<li><a href="#verify-installation">2.5. Verify the Installation</a><ul>
<li><a href="#verify-driver-installation">2.5.1. Verify the Driver Installation</a></li>
<li><a href="#compiling-examples">2.5.2. Compiling the Examples</a></li>
<li><a href="#running-binaries">2.5.3. Running Binaries</a></li>
</ul>
</li>
</ul>
</li>
<li><a href="#additional-considerations">3. Additional Considerations</a></li>
</ul>
</nav>
<nav id="search-results">
<h2>Search Results</h2>
<ol></ol>
</nav>
<script language="JavaScript" type="text/javascript" charset="utf-8" src="../common/formatting/common.min.js"></script>
<script language="JavaScript" type="text/javascript" charset="utf-8" src="../common/scripts/omniture/s_code_us_dev_aut1-nolinktrackin.js"></script>
<script language="JavaScript" type="text/javascript" charset="utf-8" src="../common/scripts/omniture/omniture.js"></script>
<noscript><a href="http://www.omniture.com" title="Web Analytics"><img src="http://omniture.nvidia.com/b/ss/nvidiacudadocs/1/H.17--NS/0" height="1" width="1" border="0" alt=""></img></a></noscript>
<script language="JavaScript" type="text/javascript" charset="utf-8" src="../common/scripts/google-analytics/google-analytics-write.js"></script>
<script language="JavaScript" type="text/javascript" charset="utf-8" src="../common/scripts/google-analytics/google-analytics-tracker.js"></script>
</body>
</html>
