Skip wasting time while trying to get TensorFlow-GPU up and running by stepping through the complex procedure to learn what steps are pivotal and which aren’t. Gain a basic overview of TensorFlow-GPU and why it might be the right choice for your machine learning and deep learning development environment. Then look at which version of Python will fit best for your needs and how to get it to interact properly with your TF-GPU. You’ll also find out how to gauge whether or not your graphics card is well suited to the task and what your options are based on your hardware.

Once all the basic requirements are met, we’ll install the Cuda toolkit to provide a development environment for creating high-performance GPU-accelerated applications. The toolkit includes GPU-accelerated libraries, debugging and optimization tools, a C/C++ compiler, and a runtime library to deploy your applications.

We’ll also need to install Visual Studio IDE for the C++ development libraries that will be required by the toolkit. Many users miss this step and they run into the problem of their toolkit not installing properly.

Having done all these steps we will then look into cuDNN, which is a deep neural network library. This library provides highly tuned implementations for standard routines such as forward and backward convolution, pooling, normalization, and activation layers. We’ll have to provide a path to cuDNN, too, as it’s not an application but a library. So we’ll look into how to make our system locate these libraries.

Finally, we’ll install TensorFlow-GPU. We’ll verify it by running some basic commands and also verify whether it’s making use of your GPU or not. TensorFlow-GPU offers a powerful, hardware-dependent development environment for the most consumptive of deep learning tasks.

What You Will Learn

Install the packages needed for TensorFlow-GPU
Set a path in Windows 10
Use the correct version of Python for your needs
Who This Video Is For

Developers new to deep learning who would like to use the more powerful GPU for hardware intensive deep-learning applications.