Work out instructions per second instantly with clear inputs, formula shown and shareable results.
Instruction throughput is clock times IPC times cores. Utilisation matters because real workloads rarely keep every core busy — memory stalls, synchronisation and load imbalance mean 50-70% is common, so peak figures overstate delivered performance substantially.
Instruction throughput
GIPS = clock (GHz) x IPC x cores x utilisation
Instruction counts differ between architectures and compilers. A CISC instruction may do the work of several RISC ones, so MIPS says nothing about useful work done.
Application-level benchmarks that measure work completed — SPEC, or better still your own workload.