Date: Wed, 21 Mar 2018 00:00:00 +0000
<p><a href="https://twitter.com/ctnzr">Bryan Catanzaro</a>, the VP Applied Deep Learning Research at NVIDIA, joins <a href="https://twitter.com/Neurotic">Mark</a> and <a href="https://twitter.com/nyghtowl">Melanie</a> this week to discuss how his team uses applied deep learning to make NVIDIA products and processes better. We talk about parallel processing and compute with GPUs as well as his team’s research in graphics, text and audio to change how these forms of communication are created and rendered by using deep learning.</p> <p>This week we are also joined by a special co-host, <a href="https://twitter.com/ffpaladin">Sherol Chen</a> who is a developer advocate on GCP and machine learning researcher on Magenta at Google. Listen at the end of the podcast where Mark and Sherol chat about all things GDC.</p> <h5 id="bryan-catanzaro">Bryan Catanzaro</h5> <p>Bryan Catanzaro is VP of Applied Deep Learning Research at NVIDIA, where he leads a team solving problems in domains ranging from video games to chip design using deep learning. Bryan earned his PhD from Berkeley, where he focused on parallel computing, machine learning, and programming models. He earned his MS and BS from Brigham Young University, where he worked on higher radix floating-point representations for FPGAs. Bryan worked at Baidu to create next generation systems for training and deploying deep learning models for speech recognition. Before that, he was a researcher at NVIDIA, where he worked on programming models for parallel processors, as well as libraries for deep learning, which culminated in the creation of the widely used CUDNN library.</p> <h5 id="cool-things-of-the-week">Cool things of the week</h5> <ul> <li>NVIDIA Tesla V100s coming to Google Cloud <a href="https://cloudplatformonline.com/CP-2018-NA-NVIDIA.html">site</a></li> <li>Automatic Serverless Deployment with Cloud Source Repositories <a href="https://cloudplatform.googleblog.com/2018/03/automatic-serverless-deployments-with-Cloud-Source-Repositories-and-Container-Builder.html"> blog</a></li> <li>Magenta <a href="https://magenta.tensorflow.org/">site</a> <ul> <li>NSynth Super <a href="https://magenta.tensorflow.org/nsynth-super">site</a></li> <li>MusicVAE <a href="https://magenta.tensorflow.org/music-vae">site</a></li> <li>Making music using new sounds generated with machine learnnig <a href="https://www.blog.google/topics/machine-learning/making-music-using-new-sounds-generated-machine-learning/"> blog</a></li> </ul> </li> <li>Building Blocks of Interpretability <a href="https://distill.pub/2018/building-blocks/">blog</a></li> </ul> <h5 id="interview">Interview</h5> <ul> <li>NVIDIA <a href="http://www.nvidia.com/page/home.html">site</a></li> <li>NVIDIA GPU Technology Conference (GTC) <a href="https://www.nvidia.com/en-us/gtc/">site</a></li> <li>CUDA <a href="https://developer.nvidia.com/cuda-zone">site</a></li> <li>cuDNN <a href="https://developer.nvidia.com/cudnn">site</a></li> <li>NVIDIA Volta <a href="https://www.nvidia.com/en-us/data-center/volta-gpu-architecture/">site</a></li> <li>NVIDIA Tesla P4 <a href="http://images.nvidia.com/content/pdf/tesla/Tesla-P4-Product-Brief.pdf"> docs</a></li> <li>NVIDIA Tesla V100s <a href="http://www.nvidia.com/page/home.html">site</a></li> <li>Silicon Valley AI Lab Baidu Research <a href="http://research.baidu.com/">site</a></li> <li>ICML: International Conference on Machine Learning <a href="https://icml.cc/">site</a></li> <li>CVPR: Computer Vision and Pattern Recognition Conference <a href="http://cvpr2018.thecvf.com/">site</a></li> </ul> <p>Referenced Papers & Research:</p> <ul> <li>Deep learning with COTS HPC System <a href="http://ai.stanford.edu/~acoates/papers/CoatesHuvalWangWuNgCatanzaro_icml2013.pdf"> paper</a></li> <li>Building High-level Features Using Large Scale Unsupervised Learning <a href="https://arxiv.org/pdf/1112.6209.pdf">paper</a></li> <li>OpenAI Learning to Generate Reviews and Discovering Sentiment <a href="https://arxiv.org/pdf/1704.01444.pdf">paper</a></li> <li>Progressive Growing of GANs for Improved Quality, Stability, and Variation <a href="https://arxiv.org/pdf/1710.10196.pdf">paper</a> and CelebA <a href="http://mmlab.ie.cuhk.edu.hk/projects/CelebA.html">dataset</a></li> <li>High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANs <a href="https://arxiv.org/pdf/1711.11585.pdf">paper</a></li> <li>Deep Image Prior <a href="https://dmitryulyanov.github.io/deep_image_prior">site</a></li> <li>How a Japanese cucumber farmer is using deep learning and TensorFlow <a href="https://cloud.google.com/blog/big-data/2016/08/how-a-japanese-cucumber-farmer-is-using-deep-learning-and-tensorflow"> blog</a></li> </ul> <p>Sample Talks:</p> <ul> <li>Future of AI Hardware Panel <a href="https://vimeo.com/238818665">video</a></li> <li>High Performance Computing is Supercharging AI <a href="https://blogs.nvidia.com/blog/author/baiduguestauthor/">blog/video</a></li> <li>AI Podcast: Where is Deep Learning Going Next? <a href="https://blogs.nvidi.com/blog/2016/12/07/ai-podcast-deep-learning-going-next/"> blog/video</a></li> </ul> <p>Sample Resources:</p> <ul> <li>Coursera How Google does Machine Learning <a href="https://www.coursera.org/learn/google-machine-learning">site</a></li> <li>NVIDIA Deep Learning Institute <a href="https://www.nvidia.com/en-us/deep-learning-ai/education/">site</a></li> <li>Udacity AI Nanodegree <a href="https://www.udacity.com/course/artificial-intelligence-nanodegree--nd889"> site</a></li> <li>Kaggle <a href="https://www.kaggle.com/">site</a></li> <li>TensorFlow <a href="https://www.tensorflow.org/">site</a></li> <li>PyTorch <a href="http://pytorch.org/">site</a></li> <li>Keras <a href="https://keras.io/">site</a></li> </ul> <div style="text-align: center;"><a href="http://www.nvidia.com/page/home.html"><img src="https://googlecloudpodcast.libsyn.com/images/post/nvidia2.png" style="margin: auto;" /></a></div> <h5 id="question-of-the-week">Question of the week</h5> <p>What to watch out for and get involved in at the Game Developers Conference (GDC) this year and in the future?</p> <ul> <li>International Grame Developers Association (IGDA) <a href="https://www.igda.org/">site</a></li> <li>Fellowship of GDC Parties <a href="https://www.facebook.com/groups/TheFellowshipOfParties/about/">site</a></li> <li>ALtCtrlGDC <a href="http://www.gdconf.com/events/altctrlgdc.html">site</a></li> <li>Experimental Gameplay Workshop <a href="http://www.experimental-gameplay.org/">site</a></li> <li>Women in Games International (WIGI) <a href="https://getwigi.com/">site</a></li> <li>Blacks in Gaming (BIG) <a href="https://www.facebook.com/Blacks-in-Gaming-142639569140671/">site</a></li> <li>Serious Games (SIGs) <a href="http://www.igda.org/group/serious">site</a></li> <li>What’s New in Firebase and Google Cloud Platform for Games <a href="https://en.wikipedia.org/wiki/Final_Fantasy_(video_game)">site</a></li> <li>Summits to Checkout: <ul> <li>AI Game Developers Summit <a href="http://www.gdconf.com/conference/ai.html">site</a></li> <li>Game Narrative Summit <a href="http://www.gdconf.com/conference/gamenarrative.html">site</a></li> <li>Independent Games Summit <a href="http://www.gdconf.com/conference/igs.html">site</a></li> </ul> </li> <li>Additional Advice: <ul> <li>The first two days are summits which are great because topic focused</li> <li>Expo floor takes a good hour to get through</li> <li>WIGI, BIG and SIGs (Google and Microsoft) have the best food</li> <li>GDC is composed of various communities</li> <li>Bring business cards</li> <li>Check out post-mortems</li> </ul> </li> <li>Favorite Games: <ul> <li>Mass Effect <a href="https://www.masseffect.com/en-gb">site</a></li> <li>Final Fantasy <a href="https://en.wikipedia.org/wiki/Final_Fantasy_(video_game)">wiki</a></li> </ul> </li> <li>Games Mark & Sherol are currently playing: <ul> <li>Hearthstone <a href="https://playhearthstone.com/en-gb/">site</a></li> <li>Dragon Age Origins <a href="https://en.wikipedia.org/wiki/Dragon_Age:_Origins">wiki</a></li> </ul> </li> </ul> <h5 id="where-can-you-find-us-next">Where can you find us next?</h5> <p>Mark and Sherol are at the <a href="http://www.gdconf.com/">Game Developer’s Conference (GDC)</a>. You can find them via the Google at GDC 2018 <a href="https://events.withgoogle.com/google-gdc-2018/">site</a>.</p> <p>Sherol will be at <a href="https://www.tensorflow.org/dev-summit/">TensorFlow Dev Summit</a> speaking about machine learning research and creativity next week.</p>