A performance study of the Acorn RISC machine
Thesis DisciplineComputer Science
Degree GrantorUniversity of Canterbury
Degree NameMaster of Science
The design decisions behind the development of the Acorn RISC Machine (ARM) are investigated, by implementing the architecture with a software emulator, to record the effectiveness of the unusual architectural features that make the ARM architecture unique. The adaption of an existing compiler construction tool (the Amsterdam Compiler Kit) has demonstrated that an optimising compiler can exploit the RISC architecture to maximize CPU performance. By compiling high level language algorithms, a complete picture of the effectiveness of the ARM architecture to support high performance computing is formed.