Google has announced that its cloud computing service is adding support for more Nvidia GPUs. First, Nvidia K80 GPUs are now generally available on Google Compute Engine. Second, Nvidia P100 GPUs are now available in beta. These high-performance GPUs are particularly helpful for enterprises running artificial intelligence, machine learning and advanced analytics workloads.
In addition, Google now has a sustained pricing model for the GPU instances. This means that organizations the use the GPUs for sustained periods of time can get up to a 30 percent discount on their cloud computing bills.
The GPUs can be used in virtual machines (VMs) or containers, and they are available in four zones: US West: Oregon, US East: South Carolina, Europe West: Belgium and Asia East: Taiwan.