Predictable CPU Architecture Designed for Small Real-Time Application - Concept and Theory of Operation

The purpose of this paper is to describe an predictable CPU architecture, based on the five stage pipeline assembly line and a hardware scheduler engine. We aim at developing a fine-grained multithreading implementation, named nMPRA-MT. The new proposed architecture uses replication and remapping techniques for the program counter, the register file, and the pipeline registers and is implemented with a FPGA device. An original implementation of a MIPS processor with thread interleaved pipeline is obtained, using dynamic scheduling of hard real-time tasks and interrupts. In terms of interrupts handling, the architecture uses a particular method consisting of assigning interrupts to tasks, which insures an efficient control for both the context switch, and the system real-time behavior. The originality of the approach resides in the predictability and spatial isolation of the hard real-time tasks, executed every two clock cycles. The nMPRA-MT architecture is enabled by an innovative scheme of predictable scheduling algorithm, without stalling the pipeline assembly line. Keywords—fine-grained multithreading; hardware scheduler; pipeline; hard real-time; predictable