While I'm not an expert, I wanted to detail what I did to get tensorflow-gpu working with my fresh Ubuntu 18.04 LTS install. NVIDIA doesn't have any official downloads for Ubuntu 18.04 yet, but you can get things to work with the available files for Ubuntu 17.04. The first thing you should check is that you have an Nvidia driver installed for your graphics card. Your graphics card must support at least Nvidia compute 3.0 to install tensorflow-gpu. You can check what graphics driver you have installed with thenvidia-smicommand.

It's only been two months since the last release from TensorFlow. However, they've certainly been busy. This update focuses mostly on bug fixes, API changes, and a few new features. Let's take a look at what's new in this ML favorite. This release focuses on improved support, documentation, and a few other API changes. TensorFlow 1.6 has added a second version of their Getting Stared document, aimed specifically at ML newcomers.

The AMIs also come with improved framework support for NVIDIA Volta. They include PyTorch v0.3.0, and support NVIDIA CUDA 9 and cuDNN 7, with significant performance improvements for training models on NVIDIA Volta GPUs. As well, they include a version of TensorFlow built from the master and merged with NVIDIA processors for Volta support. We've also added Keras 2.0 support on the CUDA 9 version of the AWS Deep Learning AMIs to work with TensorFlow as the default backend.

Since version 1.2, Google dropped GPU support on macOS from TensorFlow. As of today, the last Mac that integrated an nVidia GPU was released in 2014. Only their latest operating system, macOS High Sierra, supports external GPUs via Thunderbolt 3.1 Who doesn't have the money to get one of the latest MacBook Pro, plus an external GPU enclosure, plus a GPU, has to purchase an old MacPro and fit a GPU in there. Any way you see it, it's quite a niche market. There's another community that Google forgot.

At the 2017 GPU Technology Conference NVIDIA announced CUDA 9, the latest version of CUDA's powerful parallel computing platform and programming model. In this post I'll provide an overview of the awesome new features of CUDA 9. The CUDA 9 release candidate is now available. To learn more you can watch the recording of my talk from GTC 2017, "CUDA 9 and Beyond". The soul of CUDA 9 is support for the powerful new Volta Architecture, specifically the new Tesla V100 GPU accelerator which was launched at GTC 2017.