Help is available by moving the cursor above any 
symbol or by checking MAQAO website.
| Total Time (s) | 150.87 | ||
| Max (Thread Active Time) (s) | 7.75 | ||
| Average Active Time (s) | 6.00 | ||
| Activity Ratio (%) | 4.51 | ||
| Average number of active threads | 9.548 | ||
| Affinity Stability (%) | 79.9 | ||
| GFLOPS | 254.191 | ||
| Time in analyzed loops (%) | 4.48 | ||
| Time in analyzed innermost loops (%) | 4.38 | ||
| Time in user code (%) | 39.7 | ||
| Compilation Options Score (%) | 100 | ||
| Array Access Efficiency (%) | 89.0 | ||
| Potential Speedups | | ||
| Perfect Flow Complexity | 1.00 | ||
| Perfect OpenMP/MPI/Pthread/TBB | 1.66 | ||
| Perfect OpenMP/MPI/Pthread/TBB + Perfect Load Distribution | 3.16 | ||
| No Scalar Integer | Potential Speedup | 1.00 | |
| Nb Loops to get 80% | 5 | ||
| FP Vectorised | Potential Speedup | 1.00 | |
| Nb Loops to get 80% | 4 | ||
| Fully Vectorised | Potential Speedup | 1.02 | |
| Nb Loops to get 80% | 9 | ||
| FP Arithmetic Only | Potential Speedup | 1.02 | |
| Nb Loops to get 80% | 7 | ||
| Source Object | Issue |
|---|---|
| ▼libllama.so | |
| ○llama-vocab.cpp | |
| ○hashtable.h | |
| ○hashtable_policy.h | |
| ▼libggml-cpu.so | |
| ○binary-ops.cpp | |
| ○traits.cpp | |
| ○vec.cpp | |
| ○common.h | |
| ○sgemm.cpp | |
| ○amx.cpp | |
| ○repack.cpp | |
| ○ggml-cpu.cpp | |
| ○ops.cpp | |
| ○quants.c | |
| ○ggml-cpu.c | |
| ○mmq.cpp | |
| ▼libggml-blas.so | |
| ○ggml-blas.cpp | |
| ▼libggml-base.so | |
| ○ggml.c | |
| ▼exec | |
| ○basic_string.h |
| Experiment Name | |||||
| Application | /beegfs/hackathon/users/eoseret/qaas_runs_test/176-060-7658/intel/llama.cpp/run/binaries/gcc_3/exec | ||||
| Timestamp | 2025-10-16 14:59:11 | Universal Timestamp | 1760619551 | ||
| Number of processes observed | 1 | Number of threads observed | 240 | ||
| Experiment Type | MPI; OpenMP; | ||||
| Machine | isix06.benchmarkcenter.megware.com | ||||
| Model Name | Intel(R) Xeon(R) 6972P | ||||
| Architecture | x86_64 | Micro Architecture | GRANITE_RAPIDS | ||
| Cache Size | 491520 KB | Number of Cores | 96 | ||
| OS Version | Linux 5.14.0-570.39.1.el9_6.x86_64 #1 SMP PREEMPT_DYNAMIC Thu Sep 4 05:08:52 EDT 2025 | ||||
| Architecture used during static analysis | x86_64 | Micro Architecture used during static analysis | GRANITE_RAPIDS | ||
| Frequency Driver | intel_pstate | Frequency Governor | performance | ||
| Huge Pages | always | Hyperthreading | on | ||
| Number of sockets | 2 | Number of cores per socket | 96 | ||
| Compilation Options | exec: GNU C++17 14.2.0 -march=graniterapids -mprefer-vector-width=256 -g -O3 -O3 -O3 -funroll-loops -ffast-math -fno-omit-frame-pointer -fcf-protection=none -fno-finite-math-only libggml-base.so: GNU C11 14.2.0 -march=graniterapids -mprefer-vector-width=256 -g -O3 -O3 -O3 -std=gnu11 -funroll-loops -ffast-math -fno-omit-frame-pointer -fcf-protection=none -fno-finite-math-only -fPIC libggml-blas.so: GNU C++17 14.2.0 -march=graniterapids -mprefer-vector-width=256 -g -O3 -O3 -O3 -std=gnu++17 -funroll-loops -ffast-math -fno-omit-frame-pointer -fcf-protection=none -fno-finite-math-only -fPIC libggml-cpu.so: GNU C++17 14.2.0 -march=graniterapids -mprefer-vector-width=256 -g -O3 -O3 -O3 -std=gnu++17 -funroll-loops -ffast-math -fno-omit-frame-pointer -fcf-protection=none -fno-finite-math-only -fPIC -fopenmp libllama.so: GNU C++17 14.2.0 -march=graniterapids -mprefer-vector-width=256 -g -O3 -O3 -O3 -funroll-loops -ffast-math -fno-omit-frame-pointer -fcf-protection=none -fno-finite-math-only -fPIC | ||||
| Comments | |||||
| Dataset | |
| Run Command | <executable> -m meta-llama-3.1-8b-instruct-Q8_0.gguf -t 192 -n 0 -p 512 -r 3 |
| MPI Command | mpirun -n <number_processes> |
| Number Processes | 1 |
| Number Nodes | 1 |
| Number Processes per Node | 1 |
| Filter | Not Used |
| Profile Start | Not Used |
| Profile Stop | Not Used |
| Maximal Path Number | 4 |