Help is available by moving the cursor above any symbol or by checking MAQAO website.
Total Time (s) | 12.85 | ||
Max (Thread Active Time) (s) | 12.10 | ||
Average Active Time (s) | 11.79 | ||
Activity Ratio (%) | 96.1 | ||
Average number of active threads | 88.031 | ||
Affinity Stability (%) | 97.9 | ||
Time in analyzed loops (%) | 70.0 | ||
Time in analyzed innermost loops (%) | 66.8 | ||
Time in user code (%) | 70.5 | ||
Compilation Options Score (%) | 87.5 | ||
Array Access Efficiency (%) | 63.2 | ||
Potential Speedups | ![]() | ||
Perfect Flow Complexity | 1.00 | ||
Perfect OpenMP/MPI/Pthread/TBB | 1.29 | ||
Perfect OpenMP/MPI/Pthread/TBB + Perfect Load Distribution | 1.45 | ||
No Scalar Integer | Potential Speedup | 1.00 | |
Nb Loops to get 80% | 3 | ||
FP Vectorised | Potential Speedup | 1.00 | |
Nb Loops to get 80% | 2 | ||
Fully Vectorised | Potential Speedup | 1.00 | |
Nb Loops to get 80% | 7 | ||
FP Arithmetic Only | Potential Speedup | 1.01 | |
Nb Loops to get 80% | 7 |
Source Object | Issue |
---|---|
▼libllama.so | |
○hashtable.h | -O3 or -Ofast is missing. |
○llama-sampling.cpp | -O3 or -Ofast is missing. |
○llama-vocab.cpp | -O3 or -Ofast is missing. |
○basic_string.h | -O3 or -Ofast is missing. |
○llama-kv-cache.cpp | -O3 or -Ofast is missing. |
▼libggml-cpu.so | |
○binary-ops.cpp | -O3 or -Ofast is missing. |
○repack.cpp | -O3 or -Ofast is missing. |
○quants.c | -O3 or -Ofast is missing. |
○ggml-cpu.cpp | -O3 or -Ofast is missing. |
○traits.cpp | -O3 or -Ofast is missing. |
○vec.cpp | -O3 or -Ofast is missing. |
○ggml-cpu.c | -O3 or -Ofast is missing. |
○ops.cpp | -O3 or -Ofast is missing. |
▼exec | |
○regex_executor.tcc | -O3 or -Ofast is missing. |
○sampling.cpp | -O3 or -Ofast is missing. |
▼[vdso] | |
▼ | |
○ | -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 | |
○basic_string.tcc | -O3 or -Ofast is missing. |
○ggml.c | -O3 or -Ofast is missing. |
Application | /home/eoseret/Tools/QaaS/qaas_runs/ip-172-31-47-249.ec2.internal/175-802-9624/llama.cpp/run/binaries/gcc_5/exec | ||||
Timestamp | 2025-09-16 13:49:52 | Universal Timestamp | 1758030592 | ||
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 6.1.109-118.189.amzn2023.aarch64 #1 SMP Tue Sep 10 08:58:40 UTC 2024 | ||||
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 | + [vdso]: N/A exec: GNU C++17 14.2.0 -mcpu=neoverse-v2 -mlittle-endian -mabi=lp64 -g -O2 -funroll-loops -ffast-math -fno-omit-frame-pointer -fcf-protection=none -fno-finite-math-only -fPIC libggml-base.so: GNU C11 14.2.0 -mcpu=neoverse-v2 -mlittle-endian -mabi=lp64 -g -O2 -std=gnu11 -funroll-loops -ffast-math -fno-omit-frame-pointer -fcf-protection=none -fno-finite-math-only -fPIC libggml-cpu.so: GNU C11 14.2.0 -mcpu=neoverse-v2 -mlittle-endian -mabi=lp64 -g -O2 -std=gnu11 -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 -mcpu=neoverse-v2 -mlittle-endian -mabi=lp64 -g -O2 -funroll-loops -ffast-math -fno-omit-frame-pointer -fcf-protection=none -fno-finite-math-only -fPIC |
Dataset | |
Run Command | <executable> -m meta-llama-3.1-8b-instruct-Q8_0.gguf -no-cnv -t 96 -n 512 -p "what is a LLM?" --seed 0 |
MPI Command | mpirun -n <number_processes> --bind-to none --report-bindings |
Number Processes | 1 |
Number Nodes | 1 |
Filter | Not Used |
Profile Start | Not Used |