A Philco multiprocessing system

ABSTRACT

Although the purpose of this paper is to describe the Philco Advanced System and the design and application considerations that gave rise to it, it will be necessary to review in some detail the Philco 213 System. The two systems are quite similar in concept and organization, and differ only in three respects:

1. Speed---The 213 System operates in the vicinity of one million single-address instructions per second for each of one to four processors; the objective for the Advanced System will be at least four times that speed.

2. Technology---The 213 uses present-production discrete-component circuits, and ferrite cores for both main memory and bulk memory; the Advanced System will use integrated circuits, thin-film main memory, and ferrite core bulk memory.

3. Vocabulary and Unit Organization---The Advanced System will be totally upward-program-compatible from the 213, but will contain additional facilities both in machine language and in internal organization of system units.

References

1. R. Segal, J. Maddox and P. Plano, "Performance Advances in a Transistorized Computer System," Proc. EJCC, 1958, p. 168.
2. F. J. Corbatö et al., The Compatible Time-Sharing System, Cambridge, Mass., M.I.T. Press, 1963.
3. H. S. Bright and B. F. Cheydleur, "On the Reduction of Turnaround Time," Proc. FJCC 1962, Washington, D. C., Spartan Books, 1962, pp. 161--169.
4. M. Kory and P. Berning, "The STL Integrated Operating System," Proc. ACM 19th Nat. Conf., Aug. 1964, N. Y., ACM, pp. E2.1.1-E2.1.25.
5. L. R. Lavine, "The Dynamic Rescheduling Operator System 212SYS," Proc. 13th TUG Meeting, Washington, D. C, Sept. 1963.
6. R. L. Patrick, "Measuring Performance," Datamation 10/7, July, 1964, pp. 24--26; also privately published report, "Time-Sharing, Its Uses and Abuses," 1964.
7. R. M. Fano, "The MAC System: A Progress Report," Technial Report MAC-TR-12, M.I.T., Project MAC, Cambridge, Mass., Oct. 1964.
8. Jules I. Schwartz, "Command Research Laboratory: Introduction to the Time-Sharing System," SDC SP-1722, 14 September 1964.
9. J. B. Dennis and E. L. Glaser, "The Structure of an On-line Information Processing System," to be published as a chapter of a forthcoming book, "Information System Sciences: Proceedings of the Second Congress," Washington, D. C., Spartan Books, 1965.
10. A. B. Shafritz, "The Use of Computers in Message Switching Networks," Proc. ACM 19th Ann. Conf., N. Y., ACM, 1964, pp. N2.3. 3--4. (See also remarks in papers by B. R. Jacobellis and F. S. Vigliante, ibid.)
11. A. W. Holt, "Program Organization and Record Keeping for Dynamic Storage Allocation," Information Processing 62, Amsterdam, North Holland Publishing Co., 1962.
12. B. Galler, B. Arden and F. Westervelt, private communications.
13. A. Perlis, private communications.
14. Philco Corporation, "Philco 212 Reference Manual TM-29," Willow Grove, Pa., 1963.
15. Work sponsored under contract with the Defense Communication Agency.
16. R. W. Bemer et al., "American Standard Code for Information Interchange," N. Y., Comm/ACM 6/8 Aug., 1963, pp. 442--426.
17. B. F. Cheydleur, "The X-3 System---A Design Study," internal report, Philco Computer Div., 1963.
18. H. S. Bright, "A Proposed 64-character Alphabet for Hardware Implementation of the IAL (ALGOL 58) Reference-Level Language," N. Y., Comm/ACM 2/5 May 1959, pp. 7--9.
19. These sections were written by A. Shapiro.
20. These sections were written by T. N. Timlin.
21. H. A. Kinslow, "The Time-Sharing Monitor System," Proc. FJCC, 1964, Washington, D. C., Spartan Books, 1964, pp. 443--454. (By private communication, Kinslow credited the dynamic memory mapper to his fellow worker, O. W. Johnson.)
22. W. F. Bauer, "Computer Design from the Programmer's Viewpoint," Proc. EJCC 1958, pp. 46--51. To quote from Bauer's paper, "The central idea here is that each large metropolitan area would have one or more of these super computers. The computers would handle a number of problems concurrently. Organizations would have input-output equipment installed on their own premises and would buy time on the central computer much the same way that the average householder buys power and water from utility companies. . . . The accounting procedure would reflect the customer's detailed use. . . . The supervisory control units schedule the multiple problems . . . changing priorities as required and distributing work among the processing units. . . . A very important characteristic . . . is that it is modular in construction, and large components can be added or subtracted to make up the configuration currently in demand at the particular installation. Thus the computer can grow and change to meet changing requirements."