Watson Studio now has GPU environments for notebook users.
This new feature will be available to Watson Studio standard and enterprise plan users in the Dallas region. There can be up to 4 Nvidia® K80 GPUs in one environment.
What is a GPU and why is it useful?
A GPU (Graphics Processing Unit) is commonly used for accelerating deep learning workloads and popular machine learning libraries, such as XGBoost. In many cases, it can achieve over 10x speed-up compare to common CPUs.
How to use a GPU notebook in Watson Studio
GPU environments have popular deep learning libraries pre-installed. Just create a new notebook, select the default GPU environment as runtime, and you are good to go. It is also easy to customize a GPU environment with your own set of libraries through Conda configuration. You can learn more about GPU environments in our documentation.