Amazon and Microsoft take aim at new workloads with souped up cloud offerings

Amazon and Microsoft have methodically been launching new services and adding features to EC2 and Azure, respectively, to make the platforms more competitive and a better fit for different applications.

The latest addition from Amazon is a new virtual server, or "instance," that has been customized for graphic intensive workloads such as large scale rendering, machine learning, video encoding, and other server-side workloads that require lots parallel processing power.

To make this possible, the g2.8xlarge instance has four high-performance Nvidia GPUs, and it can be used to run Nividia's CUDA parallel programming framework, OpenCL, DirectX, and OpenGL applications without making expensive up-front capital investments, Amazon said on Thursday.

The instance is priced from US$2.60 per hour. It also has 32 virtual CPUs, two times 120GB of storage and 64GB of RAM.

Microsoft, on the other hand, finally made its SSD-based Premium Storage offering generally available, the company said on Thursday.

More enterprises will be able to move demanding enterprise applications to the cloud, including workloads such as online transaction processing, big data, and data warehousing on SQL Server, MongoDB and Cassandra, according to Microsoft.

Premium Storage has been available as a preview since December. Users can configure up to 32TB of storage per virtual machine. From June 1, the service will cost from $17.92 per month. For that users get 128GB of storage and 500 IOPS per disk.

Send news tips and comments to mikael_ricknas@idg.com