Using GPUs to Scale and Speed-up Deep Learning
Learn how to use accelerated GPU hardware to overcome the scalability problem in deep learning.
Featured on: Oct 13, 2018
Training a complex deep learning model with a very large dataset can take hours, days and occasionally weeks to train. So, what is the solution? Accelerated hardware. You can use accelerated hardware such as Google’s Tensor Processing Unit (TPU) or Nvidia GPU to accelerate your convolutional neural network computations time on the Cloud. These chips are specifically designed to support the training of neural networks, as well as the use of trained networks (inference). Accelerated hardware has recently been proven to significantly reduce training time.But the problem is that your data might be sensitive and you may not feel comfortable uploading it on a public cloud, preferring to analyze it on-premise. In this case, you need to use an in-house system with GPU support. One solution is to use IBM’s Power Systems with Nvidia GPU and PowerAI. The PowerAI platform supports popular machine learning libraries and dependencies including Tensorflow, Caffe, Torch, and Theano.In this course, you l understand what GPU-based accelerated hardware is and how it can benefit your deep learning scaling needs. You l also deploy deep learning networks on GPU accelerated hardware for several problems, including the classification of images and videos.