| Name | Module | Coverage gcc_4 (%) | Coverage Excluding Loops gcc_4 (%) | Max Inclusive Time Over Threads gcc_4 (s) | Max Exclusive Time Over Threads gcc_4 (s) | Inclusive Time w.r.t. Wall Time gcc_4 (s) | Exclusive Time w.r.t. Wall Time gcc_4 (s) | Nb Threads gcc_4 | Deviation (coverage) gcc_4 | Deviation (walltime) gcc_4 | Categories gcc_4 | GFLOPS gcc_4 | Compilation Options |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ►(anonymous namespace)::tinygemm_kernel_vnni<block_q8_0, block_q8_0, float, 1, 64, 32>::apply(int, void const*, void const*, float*, int) | libggml-cpu.so | 86.79 | 0.00 | 24.73 | 0.01 | 24.14 | 0.00 | 192 | 1.81 | 0.47 | /scratch/users/amazouz/QAAS/service/Llama.cpp/sdp772511/175-793-6543/llama.cpp/build/llama.cpp/../gcc_4/bin/libggml-blas.so (%): 100.00 | 89.53 | GNU C++17 13.3.0 -march=graniterapids -mprefer-vector-width=128 -g -O3 -O3 -O3 -std=gnu++17 -funroll-loops -ffast-math -fno-omit-frame-pointer -fcf-protection=none -fno-finite-math-only -fPIC -fopenmp -fasynchronous-unwind-tables -fstack-protector-strong -... |
| ○Loop 400 - mmq.cpp:1597-1597 - libggml-cpu.so [...] | 86.79 | 86.79 | 24.73 | 24.73 | 24.14 | 24.14 | 192 | 1.81 | 0.47 | 89.52 | |||
| ○omp_get_num_procs | libgomp.so.1.0.0 | 11.08 | 11.08 | 4.07 | 4.07 | 3.08 | 3.08 | 192 | 2.03 | 0.56 | OMP (%): 100.00 | 0.74 | |
| ►quantize_row_q8_0 | libggml-cpu.so | 0.03 | 0.00 | 0.31 | 0.00 | 0.01 | 0.00 | 6 | 0.11 | 0.03 | /scratch/users/amazouz/QAAS/service/Llama.cpp/sdp772511/175-793-6543/llama.cpp/build/llama.cpp/../gcc_4/bin/libggml-blas.so (%): 100.00 | 1046.73 | GNU C11 13.3.0 -march=graniterapids -mprefer-vector-width=128 -g -O3 -O3 -O3 -std=gnu11 -funroll-loops -ffast-math -fno-omit-frame-pointer -fcf-protection=none -fno-finite-math-only -fPIC -fopenmp -fasynchronous-unwind-tables -fstack-protector-strong -fsta... |
| ○Loop 2154 - quants.c:298-321 - libggml-cpu.so [...] | 0.03 | 0.03 | 0.31 | 0.31 | 0.01 | 0.01 | 6 | 0.11 | 0.03 | 1046.73 |