Real Time Computer Complex (RTCC) in the Apollo era

Picture #01 Credit to NASA. Picture #02 Credit to IBM. Picture #03 Credit to NASA. Picture #04 Credit to IBM. Picture #05 Credit to NASA. Picture #06 Credit to NASA. Picture #07 Credit to NASA. Picture #08 Credit to NASA. Picture #09 The eight camera viewpoints on the RTCC floor plan. The "best guess" floor plan in picture 9 of the RTCC with its five IBM 260-J75 mainframes has been inferred from pictures like shown above. These pictures are photographs or stills from film footage. The Mission Control Center with its RTCC became operational in January 1965 during the uncrewed Gemini 2 mission. Back then the RTCC was equipped with five IBM 7094-II mainframes to support the missions. It became clear that the IBM 7094's were insufficient for the required data processing tasks for the Apollo missions. It was decided that the IBM 7094's had to be replaced by the more powerful IBM 360/75's. These IBM 360/75's became operational in 1968. System upgrades were performed continuously. For example the diskdrives IBM 2311 were gradually replaced by the IBM 2314 to reduce access times and increase storage capacity. Note to the reader: Since I have not been able to find any floor plan of the arrangement of the five IBM 360's in the RTCC, I have decided to draw the floor plan myself. I went through numerous iterations in which I tried to validate the drawn plan against photos or film footage. I have tried to collect all kinds of visual information as much as possible to have this iteration process converge towards a reasonable accurate depiction of the floor plan. I have also tried to get familair with the spatial aspects of the room itself, the used peripheral equipment, its dimensions, and the various IBM manuals, to discern the logic behind the arrangement of the equipment. There are still white spots though. I have not been able to identify some of the equipment cabinets. So I am still looking for a more elaborate drawing of the RTCC systems configuration. Also the location of the System Select Unit has still not been identified. My guess would be somewhere near cluster E in front off the control room. But I have not been able to verify my suspicion. The camera viewpoints as indicated on the floor plan in picture 9 are a reflection of this iteration effort in which a version of the drawing was validated against photo's. It took me some time first to identify the various camera locations in the RTCC room using visual clues like support columns, adjacent (control) rooms, windows, doors, ceiling, ceiling curtains, walls, and the equipment itself.

The RTCC was part of the Mission Control Center (MCC) and in conjunction with the Communication, Command and Telemetry System (CCATS) and the Display & Control System it provided and presented data on demand to flight controllers and their support groups. The Real Time Computer Complex (RTCC) was used for processing incoming data to provide data to be displayed on the various flight controller console screens and to compile program instructions and data to be uploaded to the guidance computer of the Apollo spacecraft. The RTCC could also provide the necessary computer power to provide in depth analysis and to provide recommendations on spacecraft trajectories for the on going mission. The RTCC consisted of five mainframes computers. One mainframe was active, a second mainframe was in dynamic standby and did the same work in parallel with the primary mainframe. A switch to the standby mainframe could be done swiftly in the case the primary would fail. A third mainframe was kept in reserve. The two other mainframes were used for simulations, to make preparations for the next upcoming Apollo mission or to process data from the scientific packages of the current mission which were deployed on the lunar surface by the astronauts. The overall MCC system configuration is depicted on this page about the MCC.

"Best guess" floor plan of the Real Time Computer Complex (RTCC)

RTCC floor plan As explained earlier, this floorplans has been inferred from available photos of the RTCC and might still contain some inaccuracies. Five systems can be distinguished (A, B, D, E and F) (letter C was not used). The core of each system was formed by a mainframe IBM 360 model J75. To this mainframe peripheral systems were connected: additional magnetic core memory units, disk drive units, tape units, line printers, card readers, teletype units, data adapters etc.. In the 1960's this mainframe IBM 360 was one of the most sophisticated computers of its time. One of those five systems was sufficient to support an Apollo mission. Two of the remaining 4 systems were used as backup. One mainframe was active, a second mainframe was in dynamic standby and did the same work in parallel with the primary mainframe. A switch to the standby mainframe could be done swiftly in case the primary would fail. A third mainframe was kept in reserve. The two other mainframes were used for simulations, to make preparations for the next upcoming Apollo mission or to process data from the scientific packages of the current mission which were deployed on the lunar surface by the astronauts. The RTCC was located at the first floor of the Operations Wing of the Mission Control Center (MCC) in Houston) (This MCC building had 3 wings: the Administrative Wing, the Lobby Wing and the Operations Wing) The Operations Wing had 3 floors. In 1960's and 1970's there were two Mission Operations and Control Rooms, one MOCR at the second floor and the other MOCR at the third floor).

Typical system configuration of an IBM 360/75 at the RTCC

IBM 360/75 system configuration Configuration of an IBM 360 / 75 computer as selected by NASA for the Apollo mission support. CPU 2075 The CPU was divided up into three units: 1. the Bus Control Unit (BCU); 2. the Instruction Preparation Unit (I-Unit); 3. the Instruction Execution Unit (E-Unit). The categories of functional units for each CPU unit are represented in the diagram. In the I-Unit the Program Status Word Register (PSW register) has been mentioned separately from the category "Registers". The 8 byte long PSW contains the 3 byte long Instruction Counter and a total of 40 bits of data fields containing all kinds of status and control information which were instrumental for operating the CPU. The CPU was hardwired and not microcoded to maximize its processing speed. CPU memory The CPU had access to up to four IBM 2365 units of 256 kbyte Processor Storage, a total 1 MByte of storage. Data were stored in eight byte words. So a total of 128 000 words could be stored in the main memory. Two IBM 2361 Large Core Storage (LCS), with a capacity of 2 MByte each could be added as a 4 Mbyte (512 000 words) extension of the Processor Storage. The LCS also acted as a buffering device for exchanging data and programs with the IBM 2314 disk drives. OS/360 and RTOS/360 The Operating System of an IBM 360 was residing on disks or tapes. The MCC needed to have real-time data produced by their mainframes. For that pupose the OS/360 needed to be adapted. IBM has developed an extension of the OS/360, the RTOS/360, the Real Time Operating System/360, to meet NASA's demands. Many features have been incorporated into this RTOS/360. Task management had to be significantly improved to have the mainframe manage many task in real-time virtually simultanuously. I/O Channels The CPU 2075 had seven I/O channels. At least one of these seven must be used as a multiplexer channel and the other ones as a socalled selector channel. The multiplexer channel was suitable to be shared by various low-speed devices like printers, card readers and teletypes. A selector channel was suitable for high speed devices like disk drives. It operated in the burst mode, meaning that it served one I/O device at the time. A selector channel could be divided up into eight subchannels, a total of 256 I/O devices could be addressed. In this diagram tape drives are connected to the selector subchannels of the IBM 2870 which also contained the multiplexer subchannels. In the diagram is shown that for Channel 1 one subchannel was used to send data to other MCC systems and one subchannel was used to exchange data with disk drives. The IBM 2701 data adapter unit provided a rapid demand response interface to the digital display (D.TV) in the MOCR and the RTCC. The bidirectional exchange of data between the CPU and other MCC systems was provided through the Multiplexer Channel at apparently a lower data rate. Nevertheless real-time acceptance and transmission of large amounts of data and control information could be accomplished through the use of the IBM 2902 Multiplex Line Adapter (MLA). The IBM Storage Channel was an attachment to the IBM 2860-2 Selector Channel. It provided the capability of high-speed data transfer from one location in storage to another location in storage. This transmission between storage areas may be within main storage or between main storage and the large core storage or within large core storage. This storage channel enabled swapping of data and programs from low speed storage to high speed main processor storage and vice versa to use the high speed CPU as efficient as possible. A selector channel could also be used to exchange data with another CPU. A typical IBM 360/J75 system arrangement ------

Paper model (1:25) of an IBM 360 Model J75 as used at the Real Time Computer Complex (RTCC)
--------- --------- --------- --------- --------- ---------

Site Map |  References |  Change History

Copyright 2020 by Sander Panhuyzen Comments and questions welcome. All pictures and drawings contained on these pages are the author's, unless otherwise noted. No unauthorized reproduction without permission.