• Developed a cycle-accurate simulator for a 5-stage pipelined MIPS processor in C++, ensuring correct architectural state at every clock cycle
  • Designed mechanisms to address control hazards using forwarding, stalling, and branch prediction
  • Boosted efficiency with 2-level branch prediction, improving performance by 20% over baseline implementations
  • Implemented Tomasulo’s algorithm with hazard handling for out-of-order execution, reducing cycle count by ~27% for compute-intensive workloads

Updated: