Generated by GPT-5-mini| System/370 | |
|---|---|
![]() Yves Tessier · CC BY-SA 4.0 · source | |
| Name | System/370 |
| Developer | IBM |
| Family | IBM mainframe |
| Released | 1970 |
| Discontinued | 1990s |
System/370 The System/370 was a family of IBM mainframe computers introduced in 1970, serving as a successor to earlier IBM mainframes and influencing computing at IBM, Honeywell, Unisys, Fujitsu, and other corporations. It shaped enterprise deployments at institutions such as the United States Department of Defense, National Aeronautics and Space Administration, Bank of America, British Airways, and AT&T, and intersected with technologies from vendors including Motorola, Intel, Texas Instruments, Siemens, and Hitachi. The system's development involved key figures and groups at Thomas J. Watson Research Center, Palmer Park, Hursley, and regional labs in Poughkeepsie, Endicott, and Bangalore.
The System/370 lineage emerged from the commercial success of earlier IBM models including the IBM System/360, the work of engineers like those in the IBM Featherbedding era and designers associated with projects at Watson Laboratories. Announced in 1970, the platform responded to market pressures from competitors such as DEC, UNIVAC, Burroughs Corporation, Control Data Corporation, and Honeywell Information Systems. Early milestones included the introduction of features influenced by research at MIT, Stanford University, and University of California, Berkeley, collaboration with standards bodies like ANSI and ISO, and procurement contracts with agencies including the Internal Revenue Service and the Federal Reserve. Over its commercial life the family evolved through releases and engineering efforts in locations tied to IBM Rochester, IBM Hursley, and international manufacturing centers in Germany, Japan, and Brazil.
The platform preserved the 32-bit architecture roots of its predecessor, maintaining binary compatibility and supporting address spaces shaped by engineering practices from Bell Labs and Xerox PARC research. Its microarchitecture adopted designs influenced by pipeline research at Stanford Linear Accelerator Center and cache strategies studied at Carnegie Mellon University. Memory management drew on concepts from the CAMAC era and paging techniques parallel to work at Princeton University and University of Manchester. I/O channel architecture paralleled peripheral subsystem approaches employed by DEC PDP series and leveraged controllers similar to those in systems from Fujitsu and Hitachi. The system interrelated with networking standards and protocols developed at ARPA, BBN Technologies, and Xerox network experiments.
The instruction set preserved backward compatibility with the IBM System/360 model, impacting compilers and languages developed at Bell Labs, Microsoft Research, Hewlett-Packard, and language communities around Fortran, COBOL, PL/I, and ALGOL. Microcode techniques echoed practices from University of Cambridge and ETH Zurich research groups. The programming model underlay operating environments used by teams at IBM Research, commercial language vendors like Rational Software, and academic projects at University of Illinois Urbana-Champaign. Security extensions and privileged-state mechanisms reflected design patterns also seen in processors from Motorola and Intel and informed standards at NIST and ECMA International.
A broad ecosystem of system software ran on the family, including IBM's own offerings and third-party products from companies like CA Technologies, BMC Software, Micro Focus, ADP, Informatica, and RPG tool vendors. Major operating environments were analogous to concepts implemented in UNIX derivatives and were influenced by system research at Bell Labs, AT&T Laboratories, and University of California, Berkeley groups who developed networking and file system ideas. Database and transaction processing systems from Oracle Corporation, IBM Db2, CICS, and IMS played central roles in enterprise workloads for banks, airlines, and government agencies such as NASA and Department of Defense contractors. Middleware and development tools from Microsoft, Sun Microsystems, and Novell shaped integration patterns, while virtualization initiatives paralleled conceptual work at MIT and commercial hypervisors from VMware and Citrix.
IBM produced numerous models and configurations, with hardware implementations reflecting semiconductor advances from firms like Intel, AMD, Texas Instruments, Motorola, and packaging techniques from NEC and Toshiba. Peripheral and storage subsystems were provided by suppliers including Seagate, IBM Storage Systems, Fujitsu, and Hitachi Data Systems. High-reliability deployments relied on technologies comparable to those in telecommunications equipment from Alcatel-Lucent, Siemens, and Ericsson. Custom ASIC and logic designs paralleled academic work at Caltech and Massachusetts Institute of Technology, while manufacturing draws on fabs associated with GlobalFoundries and legacy IBM facilities.
Performance benchmarking invoked methodologies from SPEC and workload studies referenced by institutions like IDC, Gartner, and Forrester Research. The system influenced transaction processing at financial giants such as JPMorgan Chase, Goldman Sachs, and retail operations at Walmart and Target. Its reliability and throughput set standards compared to architectures from DEC, Cray Research, and Sun Microsystems. The platform's adoption accelerated corporate data center growth, shaped procurement at ministries in United Kingdom, Canada, and Australia, and influenced educational computing curricula at Carnegie Mellon University and Georgia Institute of Technology.
Legacy support and emulation maintained continuity through products and projects from Fujitsu, Hitachi, Unisys, IBM Global Services, and independent efforts by groups at University of Cambridge and Oxford University. Emulation layers and virtualization concepts influenced modern hypervisors from VMware, Microsoft Azure, and Amazon Web Services and informed standards at IEEE and IETF. Preservation and historical study involve museums and archives such as the Computer History Museum, Science Museum, London, and collections at Smithsonian Institution and National Museum of Computing.