| ID | Module | Source Location | Source Function | Level | Max Time Over Threads (s) | Time w.r.t. Wall Time (s) | Coverage (% app. time) | Speedup if no scalar integer | Speedup if FP arith vectorized | Speedup if fully vectorized | Speedup if FP only | Number of paths | Vectorization Ratio (%) | Vector Length Use (%) | Flops (GFLOP/s) | CQA cycles | CQA cycles if no scalar integer | CQA cycles if FP arith vectorized | CQA cycles if fully vectorized | CQA cycles if FP only |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ○Loop 35 | attention-avx512 | attention.cpp:30-31 | main | Innermost | 1.58 | 1.58 | 33.26 | 1.00 | 1.00 | 11.52 | 2.25 | 1 | 0.00 | 8.10 | 6.30 | 9.00 | 9.00 | 9.00 | 0.78 | 4.00 |
| ○Loop 24 | attention-avx512 | attention.cpp:30-31 | main | Innermost | 1.53 | 1.53 | 32.42 | 1.00 | 1.00 | 11.52 | 2.25 | 1 | 0.00 | 8.10 | 6.74 | 9.00 | 9.00 | 9.00 | 0.78 | 4.00 |
| ○Loop 56 | attention-avx512 | attention.cpp:30-31 | main | Innermost | 0.39 | 0.38 | 8.13 | 1.00 | 1.00 | 11.52 | 2.25 | 1 | 0.00 | 8.10 | 6.48 | 9.00 | 9.00 | 9.00 | 0.78 | 4.00 |
| ○Loop 46 | attention-avx512 | attention.cpp:30-31 | main | Innermost | 0.39 | 0.38 | 8.13 | 1.00 | 1.00 | 11.52 | 2.25 | 1 | 0.00 | 8.10 | 6.46 | 9.00 | 9.00 | 9.00 | 0.78 | 4.00 |
| ○Loop 51 | attention-avx512 | attention.cpp:30-31 | main | Innermost | 0.38 | 0.38 | 8.03 | 1.00 | 1.00 | 11.52 | 2.25 | 1 | 0.00 | 8.10 | 6.61 | 9.00 | 9.00 | 9.00 | 0.78 | 4.00 |
| ○Loop 31 | attention-avx512 | attention.cpp:47-48 | main | Innermost | 0.08 | 0.08 | 1.58 | 1.00 | 1.00 | 16.00 | 1.00 | 1 | 0.00 | 6.25 | 1.80 | 32.00 | 32.00 | 32.00 | 2.00 | 32.00 |
| ○Loop 36 | attention-avx512 | attention.cpp:27-27,attention.cpp:31-33 | main | InBetween | 0.05 | 0.06 | 1.16 | 1.44 | 1.44 | 11.50 | 2.88 | 1 | 23.53 | 12.13 | 12.54 | 11.50 | 8.00 | 8.00 | 1.00 | 4.00 |
| ○Loop 47 | attention-avx512 | attention.cpp:27-27,attention.cpp:31-33 | main | InBetween | 0.03 | 0.02 | 0.53 | 1.02 | 1.00 | 8.17 | 8.17 | 1 | 29.63 | 13.89 | 5.04 | 8.17 | 8.00 | 8.17 | 1.00 | 1.00 |
| ○Loop 25 | attention-avx512 | attention.cpp:27-27,attention.cpp:31-33 | main | InBetween | 0.03 | 0.02 | 0.53 | 1.08 | 1.00 | 8.67 | 8.67 | 1 | 28.57 | 13.84 | 5.10 | 8.67 | 8.00 | 8.67 | 1.00 | 1.00 |
| ○Loop 57 | attention-avx512 | attention.cpp:27-27,attention.cpp:31-33 | main | InBetween | 0.02 | 0.02 | 0.42 | 1.04 | 1.00 | 8.33 | 8.33 | 1 | 30.77 | 13.94 | 5.88 | 8.33 | 8.00 | 8.33 | 1.00 | 1.00 |
| ○Loop 27 | attention-avx512 | attention.cpp:43-44,attention.cpp:47-47,attention.cpp:52-61,attention.cpp:98-98,attention.cpp:284-284 | main | InBetween | 0.02 | 0.02 | 0.42 | 1.90 | 1.87 | 9.96 | 4.38 | 72 | 38.16 | 21.83 | 9.58 | 28.50 | 15.00 | 15.25 | 2.86 | 6.50 |
| ○Loop 52 | attention-avx512 | attention.cpp:27-27,attention.cpp:31-33 | main | InBetween | 0.01 | 0.02 | 0.32 | 1.02 | 1.00 | 8.17 | 8.17 | 1 | 29.63 | 13.89 | 6.63 | 8.17 | 8.00 | 8.17 | 1.00 | 1.00 |
| ○Loop 32 | attention-avx512 | attention.cpp:26-27,attention.cpp:31-33,attention.cpp:262-262 | main | InBetween | 0.01 | 0.01 | 0.21 | 1.00 | 1.00 | 11.08 | 3.00 | 1 | 0.00 | 10.07 | 3.60 | 6.00 | 6.00 | 6.00 | 0.54 | 2.00 |
| ○Loop 28 | attention-avx512 | attention.cpp:55-56 | main | Innermost | 0.01 | 0.01 | 0.21 | 1.50 | 1.80 | 2.00 | 1.50 | 1 | 100.00 | 50.00 | 0.70 | 1.50 | 1.00 | 0.83 | 0.75 | 1.00 |
| ○Loop 48 | attention-avx512 | attention.cpp:26-27,attention.cpp:31-31,attention.cpp:193-193 | main | InBetween | 0.00 | 0.00 | 0.11 | 1.00 | 1.00 | 12.52 | 3.00 | 1 | 0.00 | 9.72 | 1.70 | 6.00 | 6.00 | 6.00 | 0.48 | 2.00 |
| ○Loop 29 | attention-avx512 | attention.cpp:52-53 | main | Innermost | 0.00 | 0.00 | 0.11 | 1.00 | 1.75 | 2.00 | 1.00 | 1 | 100.00 | 50.00 | 3.50 | 3.00 | 3.00 | 1.71 | 1.50 | 3.00 |
| ○Loop 83 | attention-avx512 | __intel_avx_rep_memset | Single | 0.00 | 0.00 | 0.11 | 1.00 | 1.00 | 2.00 | 4.00 | 1 | 100.00 | 50.00 | 0.00 | 4.00 | 4.00 | 4.00 | 2.00 | 1.00 |