Amazon Web Services (AWS) has stolen a march on Azure and Google Cloud. The market-leading cloud provider has announced new compute instances that tap into Volta GPUs. This announcement means developers can build applications with Nvidia’s latest generation silicon.
The Volta GPUs are specifically designed to deliver high-performance acceleration for applications and artificial intelligence solutions. Computation of new machine learning models takes a heavy load. Volta GPUs provide the computation power and Amazon is the first to leverage it through the cloud.
AWS services now allows organizations to use Volta GPUs to run instances with up to 8 V100 GPUs. Amazon says the ability will be limited to its datacenters in Oregon, Northern Virginia, Ireland, and Tokyo.
One Volta GPU comes with 5,120 CUDA cores and another 640 Tensor cores. This allows one chip to deliver up to 125 TFLOPS of mixed-precision floating point, 15.7 TFLOPS of single-precision floating point, and 7.8 TFLOPS of double-precision floating point.
Amazon says customers need CUDA 9 and cuDNN7 to be able to get the most from the support:
“These drivers and libraries have already been added to the newest versions of the Windows AMIs and will be included in an updated Amazon Linux AMI that is scheduled for release on November 7th. New packages are already available in our repos if you want to to install them on your existing Amazon Linux AMI.”
Cloud Competition
I have previously discussed about the cloud market and its locked position. Amazon is comfortable at the top of the market, while Microsoft is easily ahead of Google in second place. AWS faces the tough prospect of creating growth while two major competitors behind it.
It has largely been a tough task. Microsoft, especially, has gone after AWS with its Azure platform. Lat week, the company debuted a cloud map to show how both services stack up. The implication from Microsoft is that Azure offers more. However, Amazon has the lead in Volta GPU adoption.