21 Dec 2011 Nashua - NextComputing, a provider of small form-factor workstations and servers, has made a new addition to its Nucleus family of compact workstations: the Nucleus GP. The Nucleus GP is a small form-factor, mini-tower workstation optimized for GP-GPU computing (General-Purpose computation on Graphics Processing Units). This compact desktop supercomputer provides dense GPU processing, as well as traditional CPU processing, memory, and high-speed storage, into the smallest package available. This makes the Nucleus GP ideal for space-limited environments or for transporting systems among different locations such as large research facilities.
Using the GPU to accelerate complex computations, particularly those of a parallel nature, has been largely adopted become an accepted practice within academia, as well as government and commercial applications. GP-GPU computing is particularly mature in the areas of bioinformatics, medical imaging, and weather modelling, and has gained more recent widespread acceptance in computational finance, engineering simulation, intelligence encryption/decryption, and digital content creation.
Hardware used for this type of computing is usually limited to clusters of rack-mounted servers, or large desk-side tower workstations. However, there are many situations or environments where access to a cluster is not possible, and a traditional workstation is not practical. For example, research labs with multiple departments spread among different buildings may benefit from a GPU-enabled system that is more easily transported. Space-cramped offices like those sometimes found in converted lofts may benefit high-frequency trading groups. And application domain consultants working with clients around the globe may need to set up shop anywhere their services are needed. A desktop supercomputer that is easy to travel with may be essential to their businesses.
The Nucleus GP offers the same features of a high-end GPU workstation in a system that is around half the size and weight. This system features: