| Achieved Vectorization Efficiency Achieved Vectorization Efficiency = (Estimated Gain/Vector Length) * 100% Estimated Gain = Vector Length = | |
 | Grey color = Achieved vectorization efficiency is lower than reference efficiency for original scalar loop |
 | Orange color = Achieved vectorization efficiency is higher than reference efficiency for original scalar loop |
 | Efficiency is approximately , which means actual efficiency may be lower |
| This simplified formula does not account for additional speed-ups related to microarchitecture or ISA code generation |
 | Reference Efficiency for original scalar loop Reference Efficiency = (1x/Vector Length) * 100% |
 | (100%) Theoretical Maximum Vectorization Efficiency Maximum Vectorization Efficiency = (Theoretical Maximum Gain/Vector Length) * 100% Theoretical Maximum Gain = Currently selected Vector Length = |