# MIPS Architecture ## Topology SLMIPSCPU -> SLClock -> SLInstructionFetch -> SLPipeline -> SLRegisterFile -> SLExecutionUnit -> SLIOBus -> SLInterruptController -> SLBusArbiter -> SLCoupler Optional implementation layers: SLMMU -> SLTLB SLCache -> SLL1/SLL2/SLL3 ## Register datapath The register file supplies two source operands and receives a destination result. GPR 0 is hardwired to zero at the architectural boundary. ## Five-stage abstraction ### IF Fetch the fixed-width instruction and advance sequential PC state. ### ID Decode opcode/function fields, read registers, and generate immediate/control information. ### EX Perform ALU operations, address calculation, comparisons, shifts, and branch target calculation. ### MEM Perform load/store bus transactions. ### WB Commit register results. ## Hazards A concrete MIPS implementation may resolve hazards with forwarding, interlocks, stalls, or scheduling. SLeeLa records the chosen mechanism in the implementation rather than assuming all MIPS CPUs behave identically. ## Exceptions CP0 and exception control are separate from ordinary ALU execution. Exception entry saves the implementation-defined return state and transfers control through the documented exception mechanism.