CDC 3600
Generated by GPT-5-miniExpansion Funnel Raw 72 → Dedup 0 → NER 0 → Enqueued 0
| CDC 3600 | |
|---|---|
| Name | CDC 3600 |
| Manufacturer | Control Data Corporation |
| Designer | Seymour Cray |
| Introduced | 1963 |
| Discontinued | 1971 |
CDC 3600 is a third-generation mainframe computer produced by Control Data Corporation during the 1960s. It formed part of a product line that included the CDC 1604, CDC 3000 series, and CDC 6600, and it occupied a niche between scientific machines and business-oriented systems. The machine's architecture, engineering, and deployment connected it to contemporaneous developments at IBM, Honeywell, UNIVAC, DEC, and research centers such as Los Alamos National Laboratory, Lawrence Livermore National Laboratory, and NASA.
Design and Specifications
The CDC 3600 featured a 48-bit word size derived from earlier designs by Seymour Cray and Control Data Corporation teams, influenced by concepts explored at Iowa State University. Its arithmetic unit and control logic leveraged transistor logic comparable to that used in machines like the IBM 7090 and the Burroughs B5000, while memory used magnetic core technology similar to implementations at MIT and Bell Labs. The system included multiple interrupt levels inspired by research at RAND Corporation and integrated input/output channels compatible with peripheral vendors such as IBM, Teletype Corporation, and Honeywell peripherals. The instruction set supported fixed-point, floating-point, and indexed addressing conventions paralleling those in the CDC 1604 and academic machines at University of California, Berkeley.
Physical design and packaging reflected industrial practices from Northrop Corporation and Raytheon, with cabinets sized for data centers at institutions including Princeton University, Columbia University, and Stanford University Medical Center. Cooling and power systems were engineered to standards used by General Electric and Westinghouse data installations, and the machine's chassis architecture echoed modular engineering approaches taken by Bell Labs and Fairchild Semiconductor.
History and Development
Development of the CDC 3600 began within Control Data Corporation's engineering groups led by veteran designers who had worked with Seymour Cray at CDC and had ties to projects at EMI and Ferranti. The platform arose as CDC sought to bridge market segments targeted by IBM, UNIVAC, and Honeywell and to provide an upgrade path for customers migrating from the CDC 1604 and CDC 3000 series. Prototyping drew on circuit techniques refined at Texas Instruments and Motorola, and software toolchains were influenced by compilers and operating systems developed at Carnegie Mellon University and University of Illinois Urbana-Champaign.
Field trials took place in collaboration with national laboratories—Los Alamos National Laboratory and Lawrence Livermore National Laboratory—and commercial clients such as AT&T research centers and financial institutions like Bank of America. Marketing and sales strategies were coordinated with international partners in United Kingdom, France, Germany, and Japan, reflecting export patterns similar to those of IBM and Honeywell in the 1960s.
Variants and Modifications
CDC and third-party vendors produced variants tailored to scientific computation, data processing, and real-time control, echoing customization practices seen with the CDC 6600 and CDC 6400. Some systems received enhanced memory arrays inspired by work at MIT Lincoln Laboratory and were retrofitted with I/O controllers compatible with Hewlett-Packard and DEC peripherals. Engineering groups at Princeton University and Columbia University developed academic modifications for research computing, while industrial partners such as Sperry Rand and RCA supplied alternate tape and disk controllers. Specialized configurations supported applications at NASA centers including Marshall Space Flight Center and Jet Propulsion Laboratory, and defense-related adaptations mirrored integration approaches used by General Dynamics and Lockheed.
Operational Use
Operators deployed the CDC 3600 in academic computing centers, government laboratories, and corporate data centers including installations at University of Minnesota, University of Michigan, Oak Ridge National Laboratory, and Argonne National Laboratory. Typical workloads included numerical weather prediction in collaboration with National Center for Atmospheric Research, computational fluid dynamics studies linked to Pratt & Whitney, and cryptanalysis projects analogous to historical work at Government Communications Headquarters and National Security Agency. The machine ran batch-oriented operating systems influenced by software from Carnegie Mellon University and execution environments compatible with languages such as Fortran, ALGOL, and assembly dialects used at Los Alamos National Laboratory.
Training and operations drew on practices from Bell Laboratories computing sections and university computing services, with system administrators exchanging techniques at conferences hosted by organizations like the Association for Computing Machinery and the Institute of Electrical and Electronics Engineers.
Performance and Handling
Performance of the CDC 3600 placed it between midrange scientific machines and high-end supercomputers of the era, offering throughput comparable to contemporaries such as the IBM 7094 and the UNIVAC 1108. Benchmarks for linear algebra, matrix inversion, and differential equation solvers used test suites developed at Argonne National Laboratory and Los Alamos National Laboratory, often implemented in Fortran or hand-optimized assembly. Handling characteristics—power consumption, cooling needs, and floor-space requirements—mirrored installations at National Institutes of Health research centers and corporate data facilities like General Motors research labs, necessitating raised floors, dedicated HVAC provided by firms like Carrier Corporation, and procedural maintenance aligned with practices from AT&T's technical operations.
Preservation and Legacy
Preservation efforts for CDC machines have been pursued by museums and private collections including the Computer History Museum, university archives at Iowa State University, and enthusiast groups associated with Vintage Computer Festival. Documentation, schematics, and software artifacts circulate among archives that also hold materials from Seymour Cray, Control Data Corporation, and contemporaries such as IBM and DEC. The CDC 3600 influenced later designs at Cray Research and development trajectories at Control Data Corporation subsidiaries, leaving a legacy visible in the evolution of high-performance computing at institutions including Argonne National Laboratory, Los Alamos National Laboratory, and NASA.