This article was accepted into the corpus but its outbound wikilinks were never NER-processed — typical at the deepest BFS hop or when the run's entity cap was reached. No expansion funnel to show.
| ASCI Red | |
|---|---|
| Name | ASCI Red |
| Developer | Sandia National Laboratories; Intel Corporation |
| Manufacturer | Intel Corporation |
| Introduced | 1997 |
| Discontinued | 2005 |
| Type | Supercomputer |
| Cpu | Intel Pentium Pro-derived Intel Itanium? |
| Os | UNIX System V-derived operating system? |
| Memory | 6 TB (peak) |
| Storage | Parallel file system |
| Performance | 1.068 TFLOPS (peak) |
| Power | >850 kW |
ASCI Red was a pioneering high-performance computing system deployed in the late 1990s. It achieved the first sustained teraflop-level performance and served strategic simulation tasks at Sandia National Laboratories and the United States Department of Energy. ASCI Red combined commercial microprocessor technology, industrial-scale cooling and power infrastructure, and advanced parallel software to influence subsequent systems at Lawrence Livermore National Laboratory, Los Alamos National Laboratory, and industry partners like Intel Corporation and Cray Research.
ASCI Red operated as part of the Accelerated Strategic Computing Initiative program managed by National Nuclear Security Administration and the United States Department of Energy national laboratory complex, including Sandia National Laboratories, Lawrence Livermore National Laboratory, and Los Alamos National Laboratory. The system used commodity-processor scaling strategies influenced by work at Intel Corporation, IBM research groups, and projects such as DARPA initiatives. ASCI Red's deployment embodied collaborations among Tera Computer Company, SGI, Fujitsu, and regional utilities near Albuquerque, New Mexico.
Development stemmed from directives from the United States Congress and program goals set by the National Nuclear Security Administration to maintain simulation capability after cessation of Comprehensive Nuclear-Test-Ban Treaty era testing policies debated alongside START II era negotiations. Initial proposals drew on architectures explored at University of Illinois Urbana-Champaign, Massachusetts Institute of Technology, Stanford University, and industrial labs at Intel Corporation and Cray Research. Prime contract and systems engineering involved Sandia National Laboratories with procurement and integration work performed by Intel Corporation and subcontractors including Seagate Technology and Sun Microsystems. Key figures in project oversight interacted with officials from Office of Science and Technology Policy and advisory panels chaired by experts from Los Alamos National Laboratory and Lawrence Livermore National Laboratory.
ASCI Red's hardware architecture used massively parallel arrays of x86-derived processors and large-scale interconnects influenced by designs from Intel Corporation and network research at Bell Labs. The machine's cabinetized racks, redundant power topology and environmental controls required coordination with Western Area Power Administration and local Public Service Company of New Mexico infrastructure. Storage subsystems leveraged technologies from Seagate Technology and file-system concepts researched at Carnegie Mellon University and University of California, Berkeley projects. Cooling and facility engineering referenced practices from General Electric industrial systems and standards from American Society of Mechanical Engineers. Hardware validation and benchmarking engaged teams with experience from National Center for Supercomputing Applications and Argonne National Laboratory.
In late 1997 and 1998 ASCI Red achieved sustained performance exceeding one teraflop, a milestone contemporaneous with records tracked by the TOP500 list maintained by Jack Dongarra and colleagues. Performance records compared to systems at Lawrence Livermore National Laboratory and Los Alamos National Laboratory and were publicized in contexts overlapping coverage by The New York Times, Scientific American, and technical venues such as ACM and IEEE. Benchmarking efforts referenced LINPACK kernels developed in collaboration with researchers at University of Tennessee and Oak Ridge National Laboratory. ASCI Red's ranking influenced procurement decisions at IBM, Cray Research, Hewlett-Packard, Fujitsu, NEC Corporation, and Hitachi.
System software and parallel libraries used ideas from Message Passing Interface standards developed by the MPI Forum with contributors from Argonne National Laboratory and University of Tennessee teams. Scientific codes executed on ASCI Red included hydrodynamics codes and finite-element packages whose development involved researchers from Los Alamos National Laboratory, Lawrence Livermore National Laboratory, Sandia National Laboratories, Princeton University, Columbia University, and California Institute of Technology. Workflow and visualization drew on tools from Lawrence Berkeley National Laboratory and collaborations with NASA centers such as NASA Ames Research Center. ASCI Red supported multidisciplinary projects involving scientists from Rutgers University, University of Michigan, University of California, Los Angeles, and industry partners like Microsoft for development environments and Oracle Corporation for data management concepts.
Operational control resided at Sandia National Laboratories with system administration practices informed by procedures from Los Alamos National Laboratory and Lawrence Livermore National Laboratory. ASCI Red ran production campaigns for scenario modeling coordinated with the National Nuclear Security Administration policy offices and scientific advisory boards including members from American Physical Society and National Academy of Sciences. Over its operational life, maintenance involved contractors such as Intel Corporation services and facility upgrades approved by New Mexico Public Regulation Commission. The system was phased out and decommissioned as successor machines at Oak Ridge National Laboratory, Lawrence Livermore National Laboratory, and Los Alamos National Laboratory based on new architectures from IBM and Cray Inc. entered service; equipment disposition followed federal property regulations overseen by General Services Administration.
ASCI Red influenced design choices at Oak Ridge National Laboratory with projects like Jaguar and later exascale planning at Argonne National Laboratory and Lawrence Berkeley National Laboratory. The project's success validated commodity-processor scaling, impacting product roadmaps at Intel Corporation, AMD, and prompting software portability efforts in communities associated with ACM and IEEE Computer Society. Lessons informed procurement policies at United States Department of Energy and education curricula at institutions such as Massachusetts Institute of Technology, University of California, Berkeley, Carnegie Mellon University, and Stanford University. ASCI Red's operational doctrines and benchmark achievements are cited in histories compiled by National Museum of American History and technical retrospectives published in Communications of the ACM and IEEE Spectrum.