{"id":304,"date":"2016-08-20T15:02:31","date_gmt":"2016-08-20T15:02:31","guid":{"rendered":"http:\/\/www.rangakrish.com\/?p=304"},"modified":"2016-09-11T15:19:56","modified_gmt":"2016-09-11T15:19:56","slug":"cuda-and-mathematica","status":"publish","type":"post","link":"https:\/\/www.rangakrish.com\/index.php\/2016\/08\/20\/cuda-and-mathematica\/","title":{"rendered":"CUDA and Mathematica"},"content":{"rendered":"

Recently I purchased a high-end desktop computer for my image processing project. Since many computations tend to take several hours to execute, I wanted to accelerate the calculations to the extent possible by adding a GPU. I chose NIVIDA’s GeForce GTX 1080<\/a>\u00a0processor-based card.<\/p>\n

Although I will be using C++ for my work, because Mathematica 11<\/a> came out around the same time I bought my computer, I wanted to explore Mathematica’s support for CUDA<\/a>. Programming in\u00a0 Mathematica is a lot easier and more compact compared to C++.<\/p>\n

Mathematica introduced support for OpenCL<\/a>\u00a0and CUDA in version 8.\u00a0 To use CUDA functions, you have to import CUDALink<\/strong><\/em> package.<\/p>\n

You can check if your system has a CUDA-capable device by calling the function CUDAQ[]<\/strong><\/em>. It returns True<\/strong><\/em> if a CUDA device is available.<\/p>\n

You can then call CUDAInformation[]<\/strong><\/em> to get more details of the CUDA hardware. See the figure below<\/p>\n

\"Basic<\/a>
Basic CUDA Check<\/figcaption><\/figure>\n

The first thing I wanted to check was how much of a performance improvement the CUDA functions brought in. I performed a simple calculation involving the traditional Fold<\/strong><\/em> and Map<\/strong><\/em> operations, and then performed the same calculation using the CUDA counterparts – CUDAFold<\/strong><\/em> and CUDAMap<\/strong><\/em>. You can see the execution timings below.<\/p>\n

\"Basic<\/a>
Basic Operations<\/figcaption><\/figure>\n

CUDA shows a significant speed up.<\/p>\n

Next, I performed a sort operation on a float array, first using the regular Sort<\/strong><\/em> and then using CUDASort<\/strong><\/em>.<\/p>\n

\"Sorting\"<\/a>
Sorting<\/figcaption><\/figure>\n

Here too, you can see a significant speed improvement.<\/p>\n

As a final check, I wanted to see the performance impact on image processing operations. For this, I computed the convolution operation on a sample image, first with CUDAImageConvolve<\/strong><\/em> and the with ImageConvolve<\/strong><\/em>.<\/p>\n

\"Image<\/a>
Image Convolution<\/figcaption><\/figure>\n

There is approximately a 25% performance hit when CUDA is not used. I am surprised that the difference is small (compared to the earlier two operations).<\/p>\n

Well, that is it for the CUDA experiment this time. Mathematica has many CUDA specific functions, and you can also add your own.<\/p>\n

Go check it out!<\/p>\n","protected":false},"excerpt":{"rendered":"

Recently I purchased a high-end desktop computer for my image processing project. Since many computations tend to take several hours to execute, I wanted to accelerate the calculations to the extent possible by adding a GPU. I chose NIVIDA’s GeForce GTX 1080\u00a0processor-based card. Although I will be using C++ for my work, because Mathematica 11 […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"advanced_seo_description":"","jetpack_seo_html_title":"","jetpack_seo_noindex":false,"jetpack_post_was_ever_published":false,"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":false,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2}},"categories":[72,17],"tags":[71,43],"class_list":["post-304","post","type-post","status-publish","format-standard","hentry","category-mathematica","category-programming","tag-cuda","tag-mathematica"],"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"jetpack_shortlink":"https:\/\/wp.me\/p9OLnF-4U","jetpack-related-posts":[{"id":348,"url":"https:\/\/www.rangakrish.com\/index.php\/2016\/09\/26\/computer-vision-with-mathematica\/","url_meta":{"origin":304,"position":0},"title":"Computer Vision with Mathematica","author":"admin","date":"September 26, 2016","format":false,"excerpt":"Over the past several weeks, I have been discussing many interesting features of Mathematica. As a continuation, today, I would like to show some cool functionality in the domain of computer vision and machine learning. The function ImageIdentify[] tries to identify the object in the given image. According to the\u2026","rel":"","context":"In "Machine Learning"","block_context":{"text":"Machine Learning","link":"https:\/\/www.rangakrish.com\/index.php\/category\/machine-learning\/"},"img":{"alt_text":"A Truck","src":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/09\/truck-fig.png?resize=350%2C200","width":350,"height":200},"classes":[]},{"id":390,"url":"https:\/\/www.rangakrish.com\/index.php\/2016\/10\/18\/torch7-and-neural-networks\/","url_meta":{"origin":304,"position":1},"title":"Torch7 and Neural Networks","author":"admin","date":"October 18, 2016","format":false,"excerpt":"This week I wanted to experiment with Torch7, a popular Machine Learning framework implemented in C\/LuaJIT. Seamless CUDA support is another plus point in favour of Torch7. I downloaded and installed Torch7 and related packages, as described here. It is important to also install cunn\u00a0and cutorch\u00a0packages if you need CUDA\u2026","rel":"","context":"In "Machine Learning"","block_context":{"text":"Machine Learning","link":"https:\/\/www.rangakrish.com\/index.php\/category\/machine-learning\/"},"img":{"alt_text":"Different Layers","src":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/10\/Code-1.png?resize=350%2C200","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/10\/Code-1.png?resize=350%2C200 1x, https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/10\/Code-1.png?resize=525%2C300 1.5x"},"classes":[]},{"id":1594,"url":"https:\/\/www.rangakrish.com\/index.php\/2019\/06\/08\/using-nodejs-in-mathematica-12\/","url_meta":{"origin":304,"position":2},"title":"Using NodeJS in Mathematica 12","author":"admin","date":"June 8, 2019","format":false,"excerpt":"In an earlier article, I had described Python integration in Mathematica 12. In addition to Python, NodeJS is also supported as a default \u201cexternal\u201d language. In today\u2019s article, I will focus on NodeJS integration. By the way, NodeJS support was introduced in Mathematica 11.2. Before using NodeJS with Mathematica 12,\u2026","rel":"","context":"In "Mathematica"","block_context":{"text":"Mathematica","link":"https:\/\/www.rangakrish.com\/index.php\/category\/mathematica\/"},"img":{"alt_text":"NodeJS Session Continued","src":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2019\/06\/Session2.png?resize=350%2C200&ssl=1","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2019\/06\/Session2.png?resize=350%2C200&ssl=1 1x, https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2019\/06\/Session2.png?resize=525%2C300&ssl=1 1.5x, https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2019\/06\/Session2.png?resize=700%2C400&ssl=1 2x"},"classes":[]},{"id":285,"url":"https:\/\/www.rangakrish.com\/index.php\/2016\/07\/22\/using-julia-to-interact-with-mathematica\/","url_meta":{"origin":304,"position":3},"title":"Using Julia to Interact with Mathematica","author":"admin","date":"July 22, 2016","format":false,"excerpt":"Mathematica is a powerful environment for symbolic and numerical computation. I have been using it for many years now. In this post\u00a0I had explained how we can use Mathematica bundled with Raspberry distribution to control littleBits devices. When I saw that there is support in Julia for interacting with Mathematica,\u2026","rel":"","context":"In "Julia"","block_context":{"text":"Julia","link":"https:\/\/www.rangakrish.com\/index.php\/category\/julia\/"},"img":{"alt_text":"Julia Session","src":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/07\/Julia-1.png?resize=350%2C200","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/07\/Julia-1.png?resize=350%2C200 1x, https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/07\/Julia-1.png?resize=525%2C300 1.5x"},"classes":[]},{"id":364,"url":"https:\/\/www.rangakrish.com\/index.php\/2016\/10\/11\/cloud-computing-with-mathematica\/","url_meta":{"origin":304,"position":4},"title":"Cloud Computing with Mathematica","author":"admin","date":"October 11, 2016","format":false,"excerpt":"Mathematica provides excellent support for cloud computation, and most of the time, it is a very simple and intuitive process. Today, let us look at some examples of cloud deployment. For using Mathematica\u2019s cloud capabilities, you need an appropriate subscription. I use Mathematica Desktop, which comes with some free cloud\u2026","rel":"","context":"In "Mathematica"","block_context":{"text":"Mathematica","link":"https:\/\/www.rangakrish.com\/index.php\/category\/mathematica\/"},"img":{"alt_text":"FormFunction","src":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/10\/FormFunction1.png?resize=350%2C200","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/10\/FormFunction1.png?resize=350%2C200 1x, https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/10\/FormFunction1.png?resize=525%2C300 1.5x"},"classes":[]},{"id":2947,"url":"https:\/\/www.rangakrish.com\/index.php\/2022\/11\/24\/using-julia-from-mathematica\/","url_meta":{"origin":304,"position":5},"title":"Using Julia from Mathematica","author":"admin","date":"November 24, 2022","format":false,"excerpt":"In an earlier article, I had shown how it is possible to interact with Mathematica from Julia. In today\u2019s article, I will share the details of how to interact with Julia from within Mathematica. Why would somebody want to execute Julia code inside Mathematica? Although Mathematica is a great symbolic\u2026","rel":"","context":"In "Julia"","block_context":{"text":"Julia","link":"https:\/\/www.rangakrish.com\/index.php\/category\/julia\/"},"img":{"alt_text":"Installing Julia","src":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2022\/11\/Julia-Shell.png?resize=350%2C200&ssl=1","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2022\/11\/Julia-Shell.png?resize=350%2C200&ssl=1 1x, https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2022\/11\/Julia-Shell.png?resize=525%2C300&ssl=1 1.5x"},"classes":[]}],"_links":{"self":[{"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/posts\/304","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/comments?post=304"}],"version-history":[{"count":0,"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/posts\/304\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/media?parent=304"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/categories?post=304"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/tags?post=304"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}