Two of the leading cloud computing providers today announced significant revisions to their on-demand virtual machines, with Rackspace rolling out new higher performing SSD-based VMs, and Amazon Web Services reducing the prices on one class of its VMs and introducing a service for graphics processing.
The moves reflect the jockeying, back-and-forth nature of the cloud computing market where when one company announces a significant new feature or price drop competitors usually follow, sometimes within days or even hours of the announcement or, in the case of today, announce the news on the same day.
Rackspace, which operates the largest OpenStack-based public cloud in the market, rolled out new VMs that will be the new default instance type for customers. They deliver up to 4X more RAM and 2X more CPU than the company's previous VMs, which will still be available.
The Performance Cloud Servers use Intel Xeon E5 Processors and come in nine sizes, ranging from 1GB for $0.04 per hour to 120GB for $5.44 per hour. 2G, 4G, 8G, 15G, 30G, 60G and 90GB sizes are also available. They come with a 10G interconnect between the machines and a 40G top of rack interconnect with the rest of the data center. The company will launch the VMs at the company's Washington D.C. data center today, and will roll out in Chicago, London and Dallas later this month, and Hong Kong and Sydney next year. Rackspace says the new VMs are ideal for high input output applications, like NoSQL databases.
AWS, meanwhile, released a second-generation of its graphic processing unit-based virtual machines, which are ideal for streaming high-definition content, such as 3-D models, games and other visualisations. The G2.2XLarge instances come with a NVIDIA GRID GPU, with 15GB of RAM and 60GB of SSD storage. They run on Intel Sandy Bridge processors running at 2.6GHz, across eight virtual CPUs.
The instance allows videos or graphics to be hosted and processed in Amazon's cloud, then streamed to any type of device, be it a laptop, desktop or even mobile device through a web browser. That allows a lightweight end-point to have the capacity of a powerful computer on its back end, which can be ideal for game streaming or computer aided design (CAD) processing, for example. That's opposed to a customer having a powerful workstation at their own site to run heavyweight computational 3-D rendering.
In addition to the new GPU-based instances, AWS also reduced the prices on its second-generation on-demand standard instance prices. The m3.xlarge instance price was dropped from $0.50 to $0.45 per hour, and reserved instance prices fell by 15%. The M3.xLarge and m3.2XLarge come with 15G and 30GB of memory and four or eight virtual CPUs respectively.
The timing of the announcements is of note as well: Amazon Web Services is hosting its second annual re:Invent customer conference next week in Las Vegas. Perhaps Rackspace was attempting to get its news out before major announcements from Amazon next week.
Rackspace officials said the new high-performance VMs are a major upgrade to the company's platform that it has been working on for months.
AWS officials said they will release new features to its cloud as soon as they're available; this year alone there have already been more than 200 feature upgrades to its cloud, says Matt Wood, AWS's GM of Data Science. "The earlier we can get features into customers' hands, the sooner we get the feedback," he says. Will there be more announcements at re:Invent? "We'll just have to wait and see," he says.