Google has announced Compute Engine, a large scale cloud service based on their existing infrastructure. The announcement came on the second day of Google I/O 2012, Google’s annual conference. An infrastructure-as-a-service (IAAS) product, Compute Engine lets you run Linux virtual machines on Google’s sweeping infrastructure. According to Google’s enterprise blog, the new product offers higher scalability, stability, performance and is more affordable than other IAAS services currently being offered.
The company also revealed at Google I/O that the Compute Engine has access to 770,000 cores. The demonstration of the Compute Engine showed a genetic application running on about 600,000 cores, besides showcasing that the Institute for Systems Biology had moved to Compute Engine after noticing that the new product showed better results than their in-house system.
Compute Engine revealed at Google I/O
Compute Engine’s official product page states that Linux Virtual Machines of one, two, four or eight cores are available with 3.75 GB of RAM per virtual core. For prices, an eight-core virtual machine with 30 GB of memory and two 1770 GB local disks will cost you $1.16 per hour. Other similar products launched by Google in the past include App Engine, Google Cloud Storage and BigQuery. For storage, there are options to store data on local disks, on Google’s persistent block device or using cloud storage options. Google also offers the ability to connect virtual machines and form clusters with Google’s own network and manage Internet connectivity through firewalls.
According to a report by ExtremeTech, Google’s latest enterprise offering directly competes with Amazon EC2 and Microsoft Azure. However, the Compute Engine may not be as huge as it seems at first glance. The report states, “777,000 cores, assuming the entire Compute Engine cluster consists of eight-core CPUs, equates to 96,250 computers. This is a huge number — probably equal to the total number of servers operated by Intel, or data centers such as The Planet or Rackspace, but considering Google is estimated to have more than one million servers in total, it’s not that huge. Amazon EC2, by comparison, is estimated to have around 450,000 physical machines. Still, almost 100,000 servers on your opening day is rather impressive; Google is obviously starting as it means to go on.” In comparison, the world’s fastest supercomputer NNSA’s Sequoia has 1.6 million cores (clocking 16 petaflops).
Compute Engine is currently in the limited beta stage, and Google is offering limited previews of the Engine.