Help is available by moving the cursor above any 
symbol or by checking MAQAO website.
| Total Time (s) | 7.70 | ||
| Max (Thread Active Time) (s) | 5.23 | ||
| Average Active Time (s) | 4.27 | ||
| Activity Ratio (%) | 75.3 | ||
| Average number of active threads | 53.213 | ||
| Affinity Stability (%) | 92.8 | ||
| Time in analyzed loops (%) | 63.7 | ||
| Time in analyzed innermost loops (%) | 61.1 | ||
| Time in user code (%) | 64.1 | ||
| Compilation Options Score (%) | 75.0 | ||
| Array Access Efficiency (%) | 77.9 | ||
| Potential Speedups | | ||
| Perfect Flow Complexity | 1.01 | ||
| Perfect OpenMP/MPI/Pthread/TBB | 1.31 | ||
| Perfect OpenMP/MPI/Pthread/TBB + Perfect Load Distribution | 1.89 | ||
| No Scalar Integer | Potential Speedup | 1.02 | |
| Nb Loops to get 80% | 3 | ||
| FP Vectorised | Potential Speedup | 1.00 | |
| Nb Loops to get 80% | 2 | ||
| Fully Vectorised | Potential Speedup | 1.34 | |
| Nb Loops to get 80% | 1 | ||
| FP Arithmetic Only | Potential Speedup | 1.74 | |
| Nb Loops to get 80% | 1 | ||
| Source Object | Issue |
|---|---|
| ▼libllama.so | |
| ▼ | |
| ○ | -g is missing for some functions (possibly ones added by the compiler), it is needed to have more accurate reports. Other recommended flags are: -O2/-O3, -march=(target) |
| ○ | -O2, -O3 or -Ofast is missing. |
| ○ | -mcpu=native is missing. |
| ▼exec | |
| ▼ | |
| ○ | -g is missing for some functions (possibly ones added by the compiler), it is needed to have more accurate reports. Other recommended flags are: -O2/-O3, -march=(target) |
| ○ | -O2, -O3 or -Ofast is missing. |
| ○ | -mcpu=native is missing. |
| ▼libggml-base.so | |
| ▼ggml.c | |
| ○ | -mcpu=native is missing. |
| ○ | -funroll-loops is missing. |
| ▼libggml-cpu.so | |
| ○vec.cpp | -funroll-loops is missing. |
| ○traits.cpp | -funroll-loops is missing. |
| ○kai_rhs_pack_nxk_qsi4c32pscalef16_qsu4c32s16s0.c | -funroll-loops is missing. |
| ○kai_matmul_clamp_f32_qsi8d32p4x8_qsi4c32p4x8_16x4_neon_i8mm.c | -funroll-loops is missing. |
| ○ops.cpp | -funroll-loops is missing. |
| ○kleidiai.cpp | -funroll-loops is missing. |
| ○ggml-cpu.cpp | -funroll-loops is missing. |
| ○kai_lhs_quant_pack_qsi8d32p4x8sb_f32_neon.c | -funroll-loops is missing. |
| ○binary-ops.cpp | -funroll-loops is missing. |
| ○ggml-cpu.c | -funroll-loops is missing. |
| ○quants.c | -funroll-loops is missing. |
| Application | /home/eoseret/Tools/QaaS/qaas_runs/ip-172-31-47-249.ec2.internal/176-138-2040/llama.cpp/run/base_runs/defaults/gcc/exec | ||||
| Timestamp | 2025-10-25 08:50:49 | Universal Timestamp | 1761382249 | ||
| Number of processes observed | 1 | Number of threads observed | 96 | ||
| Experiment Type | MPI; OpenMP; | ||||
| Machine | ip-172-31-47-249.ec2.internal | ||||
| Architecture | aarch64 | Micro Architecture | ARM_NEOVERSE_V2 | ||
| OS Version | Linux 6.1.155-176.282.amzn2023.aarch64 #1 SMP Tue Oct 7 15:53:00 UTC 2025 | ||||
| Architecture used during static analysis | aarch64 | Micro Architecture used during static analysis | ARM_NEOVERSE_V2 | ||
| Frequency Driver | NA | Frequency Governor | NA | ||
| Huge Pages | madvise | Hyperthreading | off | ||
| Number of sockets | 1 | Number of cores per socket | 96 | ||
| Compilation Options | exec: N/A libggml-base.so: GNU C11 14.2.0 -mlittle-endian -mabi=lp64 -g -O3 -O3 -std=gnu11 -fno-omit-frame-pointer -fcf-protection=none -fPIC libggml-cpu.so: GNU C11 14.2.0 -mcpu=neoverse-v2+crc+sve2-aes+sve2-sha3+nossbs+dotprod+i8mm+sve+nosme -mlittle-endian -mabi=lp64 -g -O3 -O3 -std=gnu11 -fno-omit-frame-pointer -fcf-protection=none -fPIC -fopenmp libllama.so: N/A | ||||
| Dataset | |
| Run Command | <executable> -m meta-llama-3.1-8b-instruct-Q4_0.gguf -t 96 -b 2048 -ub 512 -npp 0 -ntg 128 -npl 16 -c 16384 --seed 0 --output-format jsonl |
| MPI Command | mpirun -n <number_processes> --bind-to none --report-bindings |
| Number Processes | 1 |
| Number Nodes | 1 |
| Filter | Not Used |
| Profile Start | Not Used |
| Profile Stop | Not Used |