Help is available by moving the cursor above any 
symbol or by checking MAQAO website.
| Metric | r0 | r1 | r2 | r3 | |
|---|---|---|---|---|---|
| Total Time (s) | 5.16 | 4.74 | 4.98 | 3.59 | |
| Max (Thread Active Time) (s) | 5.16 | 4.74 | 4.98 | 3.58 | |
| Average Active Time (s) | 5.16 | 4.74 | 4.98 | 3.58 | |
| Activity Ratio (%) | 99.9 | 99.8 | 99.9 | 99.8 | |
| Average number of active threads | 0.999 | 0.998 | 0.999 | 0.998 | |
| Affinity Stability (%) | 99.8 | 99.7 | 99.8 | 99.7 | |
| GFLOPS | 6.795 | 6.939 | 6.339 | 9.781 | |
| Time in analyzed loops (%) | 99.6 | 95.7 | 99.2 | 99.6 | |
| Time in analyzed innermost loops (%) | 55.9 | 92.0 | 93.5 | 90.7 | |
| Time in user code (%) | 99.6 | 97.7 | 99.2 | 99.6 | |
| Compilation Options Score (%) | 66.7 | 100 | 100 | 100 | |
| Array Access Efficiency (%) | 27.1 | 75.1 | 26.0 | 50.7 | |
| Potential Speedups | | ||||
| Perfect Flow Complexity | 1.00 | 1.00 | 1.00 | 1.00 | |
| Perfect OpenMP/MPI/Pthread/TBB | 1.00 | 1.00 | 1.00 | 1.00 | |
| Perfect OpenMP/MPI/Pthread/TBB + Perfect Load Distribution | 1.00 | 1.00 | 1.00 | 1.00 | |
| No Scalar Integer | Potential Speedup | 1.03 | 1.01 | 1.01 | 1.12 |
| Nb Loops to get 80% | 4 | 2 | 3 | 4 | |
| FP Vectorised | Potential Speedup | 1.03 | 1.01 | 1.00 | 1.01 |
| Nb Loops to get 80% | 4 | 3 | 1 | 3 | |
| Fully Vectorised | Potential Speedup | 1.98 | 7.79 | 2.27 | 1.59 |
| Nb Loops to get 80% | 9 | 7 | 5 | 6 | |
| Only FP Arithmetic | Potential Speedup | 2.06 | 2.13 | 2.30 | 5.16 |
| Nb Loops to get 80% | 8 | 5 | 5 | 6 | |
| Source Object | Issue |
|---|---|
| ▼attention-avx512 | |
| ▼attention.cpp | |
| ○ | -march=(target) is missing. |
Enable log scale
Enable log scale
Enable log scale
Enable log scale
| r0 | r1 | r2 | r3 | |
|---|---|---|---|---|
| Experiment Name | ||||
| Application | ./attention-avx512 | same as r0 | same as r0 | same as r0 |
| Timestamp | 2026-05-13 14:06:46 | 2026-05-21 15:36:42 | 2026-05-21 15:37:29 | 2026-05-13 14:25:51 |
| Experiment Type | Sequential | same as r0 | same as r0 | same as r0 |
| Machine | isix06.benchmarkcenter.megware.com | same as r0 | same as r0 | same as r0 |
| Architecture | x86_64 | same as r0 | same as r0 | same as r0 |
| Micro Architecture | GRANITE_RAPIDS | same as r0 | same as r0 | same as r0 |
| Model Name | Intel(R) Xeon(R) 6972P | same as r0 | same as r0 | same as r0 |
| Cache Size | 491520 KB | same as r0 | same as r0 | same as r0 |
| Number of Cores | 96 | same as r0 | same as r0 | same as r0 |
| Maximal Frequency | 3.90 GHz | same as r0 | same as r0 | same as r0 |
| OS Version | Linux 5.14.0-611.16.1.el9_7.x86_64 #1 SMP PREEMPT_DYNAMIC Mon Dec 22 03:40:39 EST 2025 | Linux 5.14.0-611.54.6.el9_7.x86_64 #1 SMP PREEMPT_DYNAMIC Fri May 15 04:23:18 EDT 2026 | same as r1 | same as r0 |
| Architecture used during static analysis | x86_64 | same as r0 | same as r0 | same as r0 |
| Micro Architecture used during static analysis | GRANITE_RAPIDS | same as r0 | same as r0 | same as r0 |
| Compilation Options | attention-avx512: --driver-mode=g++ --intel -O3 -g -fno-omit-frame-pointer -grecord-command-line -mavx512f -mavx512bw -mavx512dq -mavx512vl -mfma -Rpass=loop-vectorize -Rpass-analysis=loop-vectorize -fapprox-func -fno-math-errno -ffinite-math-only -fassociative-math -fno-signed-zeros attention.cpp -o attention-avx512 -fveclib=SVML -dumpdir attention-avx512- | attention-avx512: --driver-mode=g++ --intel -O3 -g -fno-omit-frame-pointer -grecord-command-line -x GRANITERAPIDS -Rpass=loop-vectorize -Rpass-analysis=loop-vectorize -fapprox-func -fno-math-errno -ffinite-math-only -fassociative-math -fno-signed-zeros attention.cpp -o attention-avx512 -fveclib=SVML -dumpdir attention-avx512- | attention-avx512: --driver-mode=g++ --intel -O3 -g -fno-omit-frame-pointer -grecord-command-line -x GRANITERAPIDS -mprefer-vector-width=512 -Rpass=loop-vectorize -Rpass-analysis=loop-vectorize -fapprox-func -fno-math-errno -ffinite-math-only -fassociative-math -fno-signed-zeros attention.cpp -o attention-avx512 -fveclib=SVML -dumpdir attention-avx512- | attention-avx512: --driver-mode=g++ --intel -O3 -g -fno-omit-frame-pointer -grecord-command-line -x GRANITERAPIDS -mprefer-vector-width=512 -qopt-report=2 -ffp-model=fast=2 attention.cpp -o attention-avx512 -fveclib=SVML -dumpdir attention-avx512- |
| Number of processes observed | 1 | same as r0 | same as r0 | same as r0 |
| Number of threads observed | 1 | same as r0 | same as r0 | same as r0 |
| Frequency Driver | intel_pstate | same as r0 | same as r0 | same as r0 |
| Frequency Governor | performance | same as r0 | same as r0 | same as r0 |
| Huge Pages | always | same as r0 | same as r0 | same as r0 |
| Hyperthreading | on | same as r0 | same as r0 | same as r0 |
| Number of sockets | 2 | same as r0 | same as r0 | same as r0 |
| Number of cores per socket | 96 | same as r0 | same as r0 | same as r0 |
| MAQAO version | 2026.0.0 | same as r0 | same as r0 | same as r0 |
| MAQAO build | c2807bda8076cbf3c73622797383dfb0369078c6::20260507-181151 | same as r0 | same as r0 | same as r0 |
| Comments | same as r0 | same as r0 | same as r0 |