NAS History
Since its inception in the mid-1980s and continuing today, the NASA Advanced Supercomputing (NAS) Division has pioneered many technologies and techniques that have become standards for integrating supercomputers into a production environment. In fact, many other successful supercomputer centers around the world have adopted the architectural and operational models first implemented here.
The NAS organization celebrated 25 years of supercomputing leadership in April 2008. Here are some highlights spanning our first quarter century:
- First to put UNIX on supercomputers.
- First to implement TCP/IP networking in a supercomputing environment.
- First to link supercomputers and workstations together to distribute computation and visualization (what is now known as client/server).
- Developed Aeronet, the first high-speed wide-area network (WAN) connecting supercomputing resources to remote customer sites.
- Developed first batch queuing system for supercomputers, NQS, which became an industry standard.
- Developed the first UNIX-based hierarchical mass storage system (MSS/NAStore).
- Developed the NAS Parallel Benchmarks (NPB), which became the industry standard for objective evaluation of parallel computing architectures. The NPB are still widely used today.
- Developed PBS, the first batch queuing software for parallel and distributed systems, which has been transferred to the commercial sector.
- Developed and contributed to the development of numerous visualization software applications that have become industry standards, including Plot3D and FAST (Flow Analysis Software Toolkit).
- First to develop and distribute software for automatic detection and visualization of 3D flow field topology, such as vortex cores, separation lines, attachment lines (FAST topology module).
- Co-created the NASA Metacenter, the first successful attempt to dynamically distribute real-user production workloads across Agency supercomputing resources at geographically distant locations.
- First to develop and apply parallel multi-view tiled-display visualization systems (hyperwall).
- Co-developed, with industry partner SGI, the world's first IRIX single-system image 256-, 512- and 1,024-processor supercomputers.
- Co-developed, with SGI, the world's first Linux-based, single-system image 512-processor supercomputer.
- Co-developed, with industry partners SGI and Intel, one of the world's largest and fastest supercomputers, Columbia, a 10,240-processor supercluster running the Linux operating system.
- Co-developed, with SGI, the first 2,048-processor shared memory environment, as part of the Columbia supercomputer.
- First to demonstrate multi-stream concurrent visualization with negligible overhead to ongoing supercomputer computations, enabling visualization of every simulation time-step in dozens of views, including remote distribution of visualization streams in real time.
- Co-developed, with industry partner SGI, the first 2,048-processor single-system image (SSI) supercomputer, an Altix 4700 running the Linux operating system.