Fpp12a Floating-Point Processor

Summary

This manual provides comprehensive information on the programming, operation, and maintenance of the Digital Equipment Corporation FPP12A Floating-Point Processor, a peripheral device designed for the PDP-8 family, Linc-8, and PDP-12 computers. The FPP12A is designed to increase the efficiency of these mini-computers by offloading complex floating-point arithmetic tasks that would otherwise be inefficiently handled by software interpreters on the CPU.

Key features and topics covered include:

• System Description: The FPP12 acts as a parallel processor with its own instruction set, program counter, and accumulator. It utilizes direct memory access (data breaks) to fetch instructions and operands, effectively "stealing" memory cycles from the host CPU to perform arithmetic operations 20 to 100 times faster than software implementations.
• Data Formats: The processor supports two primary data formats: standard 24-bit and optional 60-bit (Extended Precision Mode/EPM). It also supports a 24-bit fixed-point mode for tasks where floating-point precision is not required.
• Programming and Initialization: The manual details the Active Parameter Table (APT) used for initializing the processor, the instruction set (data reference and special instructions), and the use of index registers for address modification and loop control.
• Hardware and Maintenance: The document includes an in-depth hardware description, explaining the internal register organization (FAC, FPC, etc.), timing generators, and the register gating system. Detailed maintenance guides, flow diagrams for instruction execution, and troubleshooting procedures are provided, including how to perform a "do it yourself" test program.
• Installation and Acceptance: The final section covers site requirements, cabling, AC power hook-up for different bus architectures, and the specific diagnostic tests required for system acceptance.