Cloud Computing vs. Grid Computing

Since we announced Cascade, SingleHop's new cloud hosting platform, some of our clients have asked us how cloud computing relates to gird computing. Cloud computing and grid computing are related concepts, in that they both involve sharing computer resources and distributing work across networks. However, there are important differences between the two, and they should not be viewed as interchangeable solutions. With that in mind, here’s a quick comparison of the two.

Grid Computing

Grid computing is a method of distributing large projects  across a network of connected computers, each of which does a small amount of the overall work.

One of the largest and best-known examples of grid computing is the SETI@home project, which uses millions of home computers to analyze radio signals from space, in an attempt to find evidence of intelligent life on other planets. By using small amounts of processing power on many computers, SETI@home operates as a supercomputer, capable of performing trillions of calculations per second.

Smaller grids are often found in university laboratories and in other environments where users need to analyze very large amounts of data.

Cloud Computing

Cloud computing, on the other hand, is a method of providing resources to many different users via the Internet. The resources provided can include software, operating systems, and hosting services, and they may reside on one server or on multiple servers, depending on the size of the network. Resources are provided on demand to each user, and users have no need to know which server they are using or how many other people are accessing the network.

The key concepts behind cloud computing are that resources are stored on computers other than the user’s own computer, and that the network has the flexibility to provide additional resources to users on an as-needed basis. More than anything else, this flexibility is what differentiates grid computing from cloud computing. A grid network may allow a single user or process to consume all available resources, creating a need for other users to wait or to reserve resources in advance, while a cloud network should always accommodate multiple requests at once as it self regulates demand and supply to all its users.

Despite the differences, cloud computing and grid computing are not mutually exclusive concepts. A grid computing network can be used for cloud computing, if it allows multiple users to access network resources on demand.