Hitachi, ViON Team for HPC

Hitachi America and ViON are working together to create solutions for HPC on Hitachi's BladeSymphony server platform

November 13, 2007

1 Min Read
NetworkComputing logo in a gray background | NetworkComputing

BRISBANE, Calif. -- Hitachi America, Ltd., a subsidiary of Hitachi, Ltd. (NYSE:HIT), and ViON Corporation today announced that they are working together to create solutions for High Performance Computing (HPC) on Hitachis BladeSymphony server platform. The companies will be collaborating on solutions targeted to address the needs of customers in several industry vertical markets, initially the higher education, general research and scientific communities, as well as the Federal Government.

One example of a HPC solution currently under development between the two companies is a virtual HPC cluster solution. This solution combines technologies and expertise from both companies to create a scalable virtualized cluster computing solution. A fundamental component of the virtual HPC cluster solution is Hitachi’s industry-first firmware-embedded virtualization feature - Virtage.

Virtage, based upon Hitachi's 40-year mainframe legacy, allows single or multiple physical server modules to be configured as a distributed cluster of virtual nodes. The virtual nodes perform distributed computational analysis in a Message Passing Interface (MPI) environment. The goal of this architecture is to maximize the use of scarce physical server resources.

The virtual HPC cluster solution can scale from two to “n” virtual machines (or Logical Partitions – LPARs). This virtual cluster capability is useful for prototyping, testing and evolving HPC cluster configurations with applicability for use in educational and research institutions. The virtual HPC cluster solution allows the researcher to economically create a computational environment which can be scaled-up to increase overall performance and throughput, and exactly match the dynamic requirements of the user community. Virtual and physical nodes may be combined into a multi-dimensional mesh to allow the creation of HPC clusters that are virtually unlimited in size for solving large computational problems.

Hitachi Ltd.

SUBSCRIBE TO OUR NEWSLETTER
Stay informed! Sign up to get expert advice and insight delivered direct to your inbox

You May Also Like


More Insights