Processors based on a programming language
Programming-language-based processors are described in Chap. 3 (page 73). Three examples are presented in this section. Two of the languages, FORTRAN and EULER, are algebraic languages operating on conventional data types, whereas IPL-VI is more like a conventional machine language operating on unconventional data types (i.e., list structures). A peculiar feature of IPL-VI is its conception of data as program (as well as of program as data) and the multiprogramming organization to which this led.
A command structure for complex information processing
The IPL-VI processor (Chap. 30) discussed in Part 3, Sec. 5, is an outgrowth of the IPL series of programming languages by Newell, Shaw, and Simon. The paper seriously treats both the language and the merits of casting a language in a hardware processor. IPL-VI was never implemented in hardware. (A partial IPL-V processor for the CDC 3600 was built at the Argonne National Laboratory.) A hardware processor for IPL-VI in the third generation would undoubtedly exist as an interpreter in a microprogrammed processor.
System design of a FORTRAN machine
This paper (Chap. 31) presents a way to map a software program into hardware. The machine's passes (or modes) correspond to activities one would see when compiling, loading, and executing a FORTRAN program.
BCD format is used for the arithmetic. The symbol table is simply organized and, therefore, has to be searched. A more serious approach for the actual implementation of such a machine might follow the lines of EULER (Chap. 32).
A microprogrammed implementation of EULER on IBM System 360/Model 30
This very clearly written paper describes a processor to implement an ALGOL-like language [Wirth and Weber, 1966]. An earlier processor was proposed to directly execute ALGOL [Anderson, 1961]. It is implemented using the Model 30 IBM System/360 P.microprogrammed. We include the paper both because it describes the Model 30 and because of EULER.
The P.language operates like a conventional compiler and operating system. The description presents clearly the process of compiling before execution.
The microprogramming aspects of the Model 30 are typical of other IBM System/360 models. The IBM approach to a P.microprogrammed is significantly different from that in Kampe's SD-2 (Chap. 29). In the 360 a microprogram instruction is encoded in a long word (60 to 100 bits, depending on the model) with a number of microcoded operations which can be selected in parallel. The SD-2 uses a short word, and only one operation is encoded in a single instruction.