Research and Development
With several HPC Research and Development teams and Research Labs all over the world NEC evaluates the newest trends in HPC. Our mission is to be at the forefront of HPC developments in all areas.
NEC monitors the latest trends in processor, network and storage technologies and collaborates with universities and other research centers to improve HPC solutions.
HPC Research and Development
NEC has several development teams that are dedicated to HPC. They design new HPC products for the market, giving NEC the capability to provide support in the full range of hardware and software.
The SX Architecture has evolved over the years, adapting to changes in technology. Our development department has successfully transformed the Vector computer to a Vector Engine integrated into today's LINUX clusters. A lot of work has been put into providing a seamless environment when it comes to hardware and software.
ScateFS - A global parallel filesystem
NEC actively develops a global parallel filesystem. It has evolved from the SX Series filesystem, to a global parallel filesystem for clusters and SX-Aurora TSUBASA, and with the addition of new hardware options, resilience and performance capabilities it is a filesystem for future HPC workloads.
Message Passing Interface
NEC have continuously developed and evolved the MPI implementation used for the SX series supercomputer. Now, with the SX-Aurora TSUBASA, the development continues and brings added value also to clusters in a standard x86 cluster environment.
NEC have created a resource management system to optimize the use of computer resources. Special considerations have been taken to optimise the runtime of jobs using different hierarchical resources.
NEC LXC³ - Cluster orchestrating software
LXC³ is NEC's cluster command and control stack for the LX series high performance Linux clusters. This development contains our collected cluster experience since our early cluster installations. The current software provides provisioning, monitoring and management of a cluster.
Our experience have shown that normal Linux tools sometimes are not developed with scalability in mind. In HPC we are moving to more and more nodes and larger systems where such capabilities become more important. With an in-house development team, we can react to new trends and provide our customers with appropriate solutions.
NEC LXFS - Lustre-based filesystem appliance
The development of a product for installing and maintaining Lustre based appliance have been ongoing for several years. Work covering the different versions of Lustre have been consolidated into an easy to use platform providing best practices when it comes to the complex Lustre software stack.