1. <language> A processor design which is not intended to be implemented as hardware, but which is the notional executor of a particular intermediate language (abstract machine language) used in a compiler or interpreter. An abstract machine has an instruction set, a register set and a model of memory. It may provide instructions which are closer to the language being compiled than any physical computer or it may be used to make the language implementation easier to port to other platforms. A virtual machine is an abstract machine for which an interpreter exists. Examples: ABC, Abstract Machine Notation, ALF, CAML, F-code, FP/M, Hermes, LOWL, Christmas, SDL, S-K reduction machine, SECD, Tbl, Tcode, TL0, WAM. 2. <theory> A procedure for executing a set of instructions in some formal language, possibly also taking in input data and producing output. Such abstract machines are not intended to be constructed as hardware but are used in thought experiments about computability. Examples: Finite State Machine, Turing Machine. (1995-03-13)

An abstract machine is a computer science theoretical model that allows for a detailed and precise analysis of how a computer system functions. It is analogous to a mathematical function in that it receives inputs and produces outputs based on predefined rules. Abstract machines vary from literal machines in that they are expected to perform correctly and independently of hardware. Abstract machines are “machines” because they allow step-by-step execution of programmes; they are “abstract” because they ignore many aspects of actual (hardware) machines. A typical abstract machine consists of a definition in terms of input, output, and the set of allowable operations used to turn the former into the latter. They can be used for purely theoretical reasons as well as models for real-world computer systems. In the theory of computation, abstract machines are often used in thought experiments regarding computability or to analyse the complexity of algorithms. This use of abstract machines is connected to th

<language> (Originally "CAML" - Categorical Abstract Machine Language) A version of ML by G. Huet, G. Cousineau, Ascander Suarez, Pierre Weis, Michel Mauny and others of INRIA and ENS. CAML is intermediate between LCF ML and SML [in what sense?]. It has first-class functions, {static type inference} with polymorphic types, user-defined {variant types} and product types, and pattern matching. It is built on a proprietary run-time system. The CAML V3.1 implementation added lazy and mutable data structures, a "grammar" mechanism for interfacing with the Yacc parser generator, pretty-printing tools, high-performance arbitrary-precision arithmetic, and a complete library. CAML V3 is often nicknamed "heavy CAML", because of its heavy memory and CPU requirements compared to Caml Light. in 1990 Xavier Leroy and Damien Doligez designed a new implementation called Caml Light, freeing the previous implementation from too many experimental high-level features, and more importantly, from the old Le_Lisp back-end. Following the addition of a native-code compiler and a powerful module system in 1995 and of the object and class layer in 1996, the project's name was changed to Objective Caml. ["The CAML Reference Manual", P. Weis et al, TR INRIA-ENS, 1989]. (2003-04-12)