{"id":10563,"date":"2022-11-10T15:27:33","date_gmt":"2022-11-10T20:27:33","guid":{"rendered":"https:\/\/blogs.mathworks.com\/deep-learning\/?p=10563"},"modified":"2024-02-12T20:40:24","modified_gmt":"2024-02-13T01:40:24","slug":"style-transfer-and-cloud-computing-with-multiple-gpus","status":"publish","type":"post","link":"https:\/\/blogs.mathworks.com\/deep-learning\/2022\/11\/10\/style-transfer-and-cloud-computing-with-multiple-gpus\/","title":{"rendered":"Style Transfer and Cloud Computing with Multiple GPUs"},"content":{"rendered":"The following post is from Nicholas Ide<\/a>, Product Manager at MathWorks.<\/em>\r\n
<\/h6>\r\nWe\u2019re headed to the SC22 supercomputing conference in Dallas next week. Thousands of people are expected to attend this year\u2019s Super Computing event; marking a large-scale return to in-person conferences. If you're one of those people, stop by and say hello! MathWorks will be there, representing Artificial Intelligence, High Performance Computing, and Cloud Computing.\r\n
<\/h6>\r\nAt the conference, we\u2019ll be luring people to our booth with free goodies; including Rubiks cubes, stickers, and live demos. Below I\u2019ll walk you through one of the demos we\u2019ll be showing. The new demo is an updated style transfer demo that runs on the cloud, applies AI to images captured by a web cam, uses a GPU to accelerate the underlying computationally intensive algorithm, and leverages multiple GPUs to increase the frame rate of processed results.\r\n
<\/h6>\r\nIf you\u2019ve ever wondered how you might use multiple GPUs to speed up a workflow, you should stop by. We\u2019ll show you how parallel constructs like parfeval can be used to leverage more of your CPU and GPU resources for independent tasks.\r\n
<\/h6>\r\n\"styled\r\n
<\/h6>\r\n \r\n

What is style transfer?<\/strong><\/p>\r\nWith style transfer, you can apply the stylistic appearance of one image to the scene content of a second image. To learn more about style transfer, read the documentation example Neural Style Transfer Using Deep Learning<\/a>.\r\n

<\/h6>\r\n
\"image<\/h6>\r\nFigure:<\/strong> Style transfer with deep learning\r\n
<\/h6>\r\nNow, some might argue that style transfer isn\u2019t exactly new, which is true. In fact, we presented a style transfer demo a few years back. Read more about our original demo in this blog post: MATLAB Demos at GTC: Style Transfer and Celebrity Lookalikes<\/a>.\r\n
<\/h6>\r\nWhat is new, is the acceleration of a computationally expensive demo by just leveraging more hardware with the same core code; speeding up an algorithm that is normally just a few frames per second, into a stream-able algorithm with 4 times that speed. In fact, our demo, which uses a high-end multi-GPU instance in the cloud, can process 15 frames per second.\r\n
<\/h6>\r\n \r\n

Connection to Cloud Machine<\/strong><\/p>\r\nTo run the style transfer demo, we connect to a cloud Windows machine in AWS using MathWorks Cloud Center<\/a>.\r\n

<\/h6>\r\nIf you have a MathWorks Account, a license for MATLAB, and an AWS account, you can leverage MathWorks Cloud Center to get on-demand access to Windows or Linux instances in the cloud with hardware that far exceeds what you likely have on your desktop now. Getting set up the first time is straightforward, and re-starting your instance is a breeze. The best part is that all the changes you make to the environment persist between re-starts. The one-time effort for initial set-up quicky pays dividends in re-use.\r\n
<\/h6>\r\n
\"Windows<\/h6>\r\nFigure:\u00a0<\/strong>Windows machine on the cloud\r\n
<\/h6>\r\nIf you are new to creating, managing, and accessing machines on AWS with MATLAB, see the documentation for Getting Started with Cloud Center<\/a> and Starting MATLAB on AWS Using Cloud Center<\/a>.\r\n
<\/h6>\r\n \r\n

GPU-Accelerated Computing<\/strong><\/p>\r\nWe used App Designer<\/a> to easily build a professional-looking app that provides an integrated environment to load frames, perform style transfer using deep learning, leverage one or more GPUs, and display results.\r\n

<\/h6>\r\nThe key aspects and controls of the app (starting from the bottom) are:\r\n
<\/h6>\r\n