sic

Single-Instruction Computer (SIC) Simulator

A SIC Factorial (Programming assignment due to Dr. Mayer Goldberg)

In their classical tome Computer Organization & Design, the Hardware/Software Interface, authors Patterson and Hennessy introduce the Single-Instruction Computer (SIC, pronounced "sick"). The original description of the SIC micro-architecture has no registers, a single assembly instruction, an unspecified address size, no mechanism for IO, and no clear mechanism for termination. We need to decide arbitrarily on some of these issues:

• The address space is 32/64-bit wide. This is also the smallest addressible chunk of memory.
• The single assembly language instruction is SBN, which stands for Subtract-and-Branch-if-Negative. The format of the instruction is
  
  SBN A, B, C
  
  and its semantics is as follows:
  
  if ((Memory[A] -= Memory[B]) < 0) goto C;
  
  The encoding of this command takes up three 32/64-bit words, and consists of the encoding of A, followed by the encoding of B, followed by the encoding of C.
• The execution of a SIC program terminates at the instruction SBN 0, 0, 0, which is encoded as three consecutive 32/64-bit zeros. This instruction doesn't accomplish anything useful, so it s a good choice for a termination criterion.
• Upon termination of the SIC program, the program dumps some output. The output is determined by the following conventions:
  1. If Memory[0] is non-zero, then the output is a contiguous list of unsigned integers. Otherwise it s a string of characters, each one of which comes from a contiguous address.
  2. Memory[1] contains the number of characters or signed integers that are to be printed.
  3. Memory[2] contains the location from which the list of integers or characters should be printed.

A SIC program can, therefore, output any number of integers, or a string of any length.

Programmer: I.A.Smith, February 2005