Cameroon GCE A-Level Computer Science (0795) Practice Questions - Cameroon General Certificate Of Education Advanced Level Computer Science (Syllabus 0795) Exam

Harvard architecture uses physically separate memory units and buses for instructions and data, whereas Von Neumann uses unified memory and buses for both. Von Neumann architecture utilizes separate arithmetic pipelines for integer and floating-point operations, whereas Harvard combines them. Harvard architecture requires all instructions to execute in a single clock cycle, whereas Von Neumann uses multi-cycle microcode. Von Neumann architecture uses direct-mapped caches exclusively, whereas Harvard architecture supports fully associative caches.