Hive Computing and High Performance Computing

Increasing numbers of businesses and other organizations are beginning to regard High Performance Computing as a mission critical function. As a result, they are growing increasingly dissatisfied with solutions like conventional supercomputers, distributed supercomputers and Linux clusters because they offer little more than just scalability.

What businesses and other organizations need are solutions that are scalable but are also survivable, usable, and offer a low Total Cost Of Ownership.

That is where Hive Computing enters the picture.

Survivable

As High Performance Computing becomes increasingly mission critical, the failure of calculations — and in particular long-running calculations — becomes increasingly problematic. It is becoming less and less acceptable to restart calculations that may take days or weeks to run.

A Hive solves this problem.

Should a computer on which a calculation is being performed fail for any reason, the Hive will automatically detect that event and restart that calculation. There is no need to restart the operation from scratch or for the programmer to know that anything happened.

Usable

Because a Hive takes over the responsibility for handling events like the failure of infrastructure components, developing a Hive-based application is an easier, faster, and less expensive process.

Low Total Cost of Ownership

Unlike distributed supercomputers and Linux clusters, which force organizations to bear the burden of maintaining the computers that make up a cluster, a Hive is self-maintaining. A Hive will automatically maintain and update the configurations of all the Workers that make up the Hive. This, when combined with the fact that a Hive is both more survivable and more usable than other solutions, means a Hive has a significantly lower Total Cost of Ownership.