| Loop id | Source Location | Source Function | Level | Max Thread Time / Walltime gcc_0 (%) | Exclusive Coverage gcc_0 (%) | Inclusive Coverage gcc_0 (%) | Max Exclusive Time Over Threads gcc_0 (s) | Max Inclusive Time Over Threads gcc_0 (s) | Exclusive Time w.r.t. Wall Time gcc_0 (s) | Inclusive Time w.r.t. Wall Time gcc_0 (s) | Nb Threads gcc_0 | GFLOPS gcc_0 | Vectorization Ratio (%) | Vector Length Use (%) | Speedup If No Scalar Integer | Speedup If FP Vectorized | Speedup If Fully Vectorized | Speedup If Perfect Load Balancing gcc_0 | Stride 0 | Stride 1 | Stride n | Stride Unknown | Stride Indirect | Array Access Efficiency |
|---|
| 408 | libggml-cpu.so - mmq.cpp:1597-1597 [...] | (anonymous namespace)::tinygemm_kernel_vnni<block_q8_0, block_q8_0, float, 1, 64, 32>::apply(int, void const*, void const*, float*, int) | Single | 25.21 | 15.34 | 15.34 | 10.98 | 10.98 | 4.15 | 4.15 | 183 | 85.89 | 87.67 | 85.7 | 1 | 1 | 1.07 | 2.54 | 0 | 4 | 2 | 0 | 1 | 78.57 |
| 521 | libggml-cpu.so - mmq.cpp:520-2194 [...] | void (anonymous namespace)::tinygemm_kernel_amx<block_q8_0, block_q8_0, float, 32, 0>(int, int, int, void const*, void const*, float*, int) [clone .constprop.0] | Outermost | 0.15 | 0.08 | 0.09 | 0.06 | 0.06 | 0.02 | 0.02 | 150 | 242.36 | NA | NA | NA | NA | NA | 2.47 | NA | NA | NA | NA | NA | 0.00 |
| 1663 | libggml-cpu.so - vec.h:491-491 [...] | ggml_compute_forward_flash_attn_ext | Innermost | 0.36 | 0.06 | 0.06 | 0.16 | 0.16 | 0.02 | 0.02 | 33 | 769.38 | 100 | 70 | 1 | 1 | 1.4 | 1.61 | 1 | 1 | 3 | 0 | 0 | 85.00 |
| 959 | libggml-cpu.so - vec.cpp:311-311 [...] | ggml_vec_dot_f16 | Single | 0.30 | 0.06 | 0.06 | 0.13 | 0.13 | 0.02 | 0.02 | 34 | 1790.37 | 100 | 66.67 | 1 | 1 | 1.67 | 1.5 | 0 | 0 | 4 | 0 | 0 | 75.00 |
| 417 | libggml-cpu.so - mmq.cpp:1563-2455 [...] | ggml_backend_amx_mul_mat(ggml_compute_params const*, ggml_tensor*)::{lambda(int, int)#2}::operator()(int, int) const::{lambda()#1}::operator()() const | Outermost | 0.24 | 0.05 | 0.05 | 0.10 | 0.10 | 0.01 | 0.01 | 86 | 1.23 | 43.9 | 30.64 | 3.16 | 1 | 1.27 | 3.63 | NA | NA | NA | NA | NA | 0.00 |
| 455 | libggml-cpu.so - mmq.cpp:822-1392 [...] | void parallel_for<(anonymous namespace)::convert_B_packed_format<block_q8_0, 32>(void*, block_q8_0 const*, int, int)::{lambda(int, int)#1}>(int, (anonymous namespace)::convert_B_packed_format<block_q8_0, 32>(void*, block_q8_0 const*, int,... | Innermost | 0.10 | 0.03 | 0.03 | 0.04 | 0.04 | 0.01 | 0.01 | 131 | 0.00 | 78.57 | 69.64 | 3.5 | 1 | 1.03 | 3.33 | 2 | 0 | 0 | 20 | 0 | 54.55 |
| 71 | libggml-cpu.so - ggml-cpu.c:1664-2891 [...] | ggml_graph_compute_thread | Single | 0.08 | 0.03 | 0.03 | 0.04 | 0.04 | 0.01 | 0.01 | 165 | 2.75 | 0 | 8.82 | 1 | 1 | 14.32 | 3.47 | NA | NA | NA | NA | NA | 0.00 |
| 996 | libggml-cpu.so - ops.cpp:6210-6245 [...] | ggml_compute_forward_rope_f32(ggml_compute_params const*, ggml_tensor*, bool) | Innermost | 0.08 | 0.03 | 0.03 | 0.04 | 0.04 | 0.01 | 0.01 | 192 | 588.56 | 19.61 | 9.8 | 1.14 | 1.71 | 5.33 | 4.06 | NA | NA | NA | NA | NA | 0.00 |
| 990 | libggml-cpu.so - ops.cpp:6446-6456 [...] | ggml_compute_forward_rope_f32(ggml_compute_params const*, ggml_tensor*, bool) | Innermost | 0.21 | 0.02 | 0.02 | 0.09 | 0.09 | 0.00 | 0.00 | 36 | 27.02 | 69.23 | 36.54 | 1 | 1 | 2.22 | 3.47 | 0 | 2 | 0 | 1 | 0 | 83.33 |
| 1661 | libggml-cpu.so - vec.h:375-751 [...] | ggml_compute_forward_flash_attn_ext | InBetween | 0.11 | 0.02 | 0.09 | 0.05 | 0.20 | 0.00 | 0.02 | 32 | 1129.75 | 27.6 | 11.89 | 2.24 | 3.04 | 1.4 | 2.05 | NA | NA | NA | NA | NA | 0.00 |
| 2214 | libggml-cpu.so - quants.c:298-321 [...] | quantize_row_q8_0 | Single | 1.54 | 0.01 | 0.01 | 0.67 | 0.67 | 0.00 | 0.00 | 1 | 413.37 | 60 | 28.75 | 1 | 1.4 | 2.84 | 1 | NA | NA | NA | NA | NA | 0.00 |
| 5 | libggml-cpu.so - ggml-cpu.c:3204-3204 [...] | ggml_cpu_fp32_to_fp16 | Single | 0.14 | 0.01 | 0.01 | 0.06 | 0.06 | 0.00 | 0.00 | 34 | 0.00 | 100 | 83.33 | 1 | 1 | 1 | 4.53 | 0 | 2 | 0 | 0 | 0 | 100.00 |
| 716 | libggml-cpu.so - binary-ops.cpp:18-32 [...] | ggml_compute_forward_mul | Innermost | 0.64 | 0.01 | 0.01 | 0.28 | 0.28 | 0.00 | 0.00 | 2 | 90.99 | 100 | 50 | 1 | 1.41 | 2 | 1.96 | 0 | 3 | 0 | 0 | 0 | 100.00 |
| 634 | libggml-cpu.so - binary-ops.cpp:10-32 [...] | ggml_compute_forward_add_non_quantized | Innermost | 0.45 | 0.00 | 0.00 | 0.20 | 0.20 | 0.00 | 0.00 | 4 | 124.16 | 100 | 50 | 1 | 1.41 | 2 | 3.63 | 0 | 3 | 0 | 0 | 0 | 100.00 |
| 963 | libggml-cpu.so - vec.cpp:372-372 [...] | ggml_vec_swiglu_f32 | Single | 0.37 | 0.00 | 0.00 | 0.16 | 0.16 | 0.00 | 0.00 | 1 | 6425.74 | 100 | 100 | 1.02 | 1 | 1 | 1 | 0 | 3 | 0 | 0 | 0 | 100.00 |
| 1322 | libggml-cpu.so - ops.cpp:4325-4326 | ggml_compute_forward_rms_norm | Innermost | 0.31 | 0.00 | 0.00 | 0.13 | 0.13 | 0.00 | 0.00 | 7 | 362.96 | 0 | 7.81 | 1 | 2.95 | 14 | 6.75 | 0 | 1 | 0 | 0 | 0 | 100.00 |
| 2943 | exec - sampling.cpp:125-126 | common_sampler_sample(common_sampler*, llama_context*, int, bool) | Single | 0.17 | 0.00 | 0.00 | 0.07 | 0.07 | 0.00 | 0.00 | 1 | 0.00 | 0 | 6.25 | 1.5 | 1 | 16 | 1 | 0 | 1 | 2 | 0 | 0 | 83.33 |
| 2573 | libllama.so - stl_algo.h:1594-1595 [...] | llama_token_data_array_partial_sort_inplace(llama_token_data_array*, int) | Single | 0.08 | 0.00 | 0.00 | 0.04 | 0.04 | 0.00 | 0.00 | 1 | 0.00 | 0 | 7.45 | 1.56 | 1 | 13.46 | 1 | 1.5 | 1 | 0.5 | 0 | 0 | 93.75 |
