| Loop id | Source Location | Source Function | Level | Max Thread Time / Walltime gcc_3 (%) | Exclusive Coverage gcc_3 (%) | Inclusive Coverage gcc_3 (%) | Max Exclusive Time Over Threads gcc_3 (s) | Max Inclusive Time Over Threads gcc_3 (s) | Exclusive Time w.r.t. Wall Time gcc_3 (s) | Inclusive Time w.r.t. Wall Time gcc_3 (s) | Nb Threads gcc_3 | Vectorization Ratio (%) | Vector Length Use (%) | Speedup If No Scalar Integer | Speedup If FP Vectorized | Speedup If Fully Vectorized | Speedup If Perfect Load Balancing gcc_3 | Stride 0 | Stride 1 | Stride n | Stride Unknown | Stride Indirect | Array Access Efficiency |
|---|
| 2240 | libggml-cpu.so - kai_matmul_clamp_f32_qsi8d32p4x8_qsi4c32p4x8_16x4_neon_i8mm.c:131-131 | kai_run_matmul_clamp_f32_qsi8d32p4x8_qsi4c32p4x8_16x4_neon_i8mm | Innermost | 37.29 | 84.37 | 84.37 | 1.70 | 1.70 | 2.42 | 2.42 | 95 | 28.14 | 55.37 | 1 | 1 | 1.85 | 1.05 | 0 | 0 | 5 | 0 | 0 | 75.00 |
| 820 | libggml-cpu.so - ops.cpp:6210-6245 [...] | ggml_compute_forward_rope_f32(ggml_compute_params const*, ggml_tensor*, bool) | Innermost | 1.31 | 1.67 | 1.67 | 0.06 | 0.06 | 0.05 | 0.05 | 96 | 2.22 | 21.18 | 1.66 | 1.48 | 5.01 | 1.89 | NA | NA | NA | NA | NA | 0.00 |
| 1613 | libggml-cpu.so - ops.cpp:8778-8939 [...] | ggml_compute_forward_flash_attn_ext | InBetween | 0.98 | 1.15 | 1.49 | 0.04 | 0.05 | 0.03 | 0.04 | 95 | 16.44 | 55.31 | 4.33 | 1.03 | 1.21 | 2.04 | NA | NA | NA | NA | NA | 0.00 |
| 2210 | libggml-cpu.so - kai_lhs_quant_pack_qsi8d32p4x8sb_f32_neon.c:96-262 [...] | kai_run_lhs_quant_pack_qsi8d32p4x8sb_f32_neon | Innermost | 1.09 | 0.85 | 0.85 | 0.05 | 0.05 | 0.02 | 0.02 | 63 | 76.21 | 96.75 | 1 | 1 | 1.03 | 2.02 | 0 | 0 | 6 | 0 | 0 | 75.00 |
| 794 | libggml-cpu.so - vec.cpp:390-390 [...] | ggml_vec_swiglu_f32 | Innermost | 0.66 | 0.76 | 0.76 | 0.03 | 0.03 | 0.02 | 0.02 | 85 | 84.85 | 96.97 | 1.07 | 1 | 1.02 | 1.84 | 0 | 0 | 3 | 0 | 0 | 75.00 |
| 2241 | libggml-cpu.so - kai_matmul_clamp_f32_qsi8d32p4x8_qsi4c32p4x8_16x4_neon_i8mm.c:131-131 | kai_run_matmul_clamp_f32_qsi8d32p4x8_qsi4c32p4x8_16x4_neon_i8mm | InBetween | 0.87 | 0.57 | 84.94 | 0.04 | 1.71 | 0.02 | 2.43 | 81 | 27.59 | 55.71 | 1 | 1 | 1.62 | 3.09 | NA | NA | NA | NA | NA | 0.00 |
| 1193 | libggml-cpu.so - ops.cpp:4325-4326 | ggml_compute_forward_rms_norm | Innermost | 0.66 | 0.43 | 0.43 | 0.03 | 0.03 | 0.01 | 0.01 | 76 | 17.78 | 55.56 | 1 | 1.25 | 2 | 2.9 | 0 | 0 | 0 | 2 | 0 | 50.00 |
| 431 | libggml-cpu.so - binary-ops.cpp:10-32 [...] | ggml_compute_forward_add_non_quantized | Innermost | 0.44 | 0.25 | 0.25 | 0.02 | 0.02 | 0.01 | 0.01 | 55 | 25 | 100 | 1 | 1 | 1 | 2.39 | 0 | 0 | 3 | 0 | 0 | 75.00 |
| 515 | libggml-cpu.so - binary-ops.cpp:18-32 [...] | ggml_compute_forward_mul | Innermost | 0.55 | 0.25 | 0.25 | 0.03 | 0.03 | 0.01 | 0.01 | 56 | 25 | 100 | 1 | 1 | 1 | 3.11 | 0 | 0 | 3 | 0 | 0 | 75.00 |
| 788 | libggml-cpu.so - vec.cpp:311-311 [...] | ggml_vec_dot_f16 | Single | 0.33 | 0.22 | 0.22 | 0.02 | 0.02 | 0.01 | 0.01 | 54 | 44.44 | 100 | 1 | 1 | 1 | 2.03 | 0 | 0 | 4 | 6 | 0 | 60.00 |
| 1622 | libggml-cpu.so - vec.h:491-491 [...] | ggml_compute_forward_flash_attn_ext | Innermost | 0.33 | 0.22 | 0.22 | 0.02 | 0.02 | 0.01 | 0.01 | 49 | 27.59 | 100 | 1 | 1 | 1 | 1.84 | 1 | 0 | 4 | 4 | 0 | 66.67 |
| 2113 | libggml-cpu.so - quants.c:2506-2575 [...] | ggml_vec_dot_q6_K_q8_K | Innermost | 0.11 | 0.19 | 0.19 | 0.01 | 0.01 | 0.01 | 0.01 | 71 | 49.44 | 66.92 | 1.06 | 1 | 1.26 | 1 | 3 | 4 | 2 | 0 | 0 | 94.44 |
| 1 | libggml-cpu.so - ggml-cpu.c:3228-3229 [...] | ggml_cpu_fp32_to_fp16 | Single | 0.44 | 0.18 | 0.18 | 0.02 | 0.02 | 0.01 | 0.01 | 45 | 86.21 | 94.83 | 1 | 1 | 1.04 | 2.69 | 2 | 0 | 0 | 0 | 0 | 100.00 |
| 2223 | libggml-cpu.so - kai_rhs_pack_nxk_qsi4c32pscalef16_qsu4c32s16s0.c:127-139 [...] | kai_run_rhs_pack_nxk_qsi4c32pscalef16_qsu4c32s16s0 | Innermost | 6.34 | 0.16 | 0.16 | 0.29 | 0.29 | 0.00 | 0.00 | 1 | 0 | 50 | 1 | 1 | 2 | 1 | 0 | 0 | 1 | 1 | 0 | 62.50 |
| 2219 | libggml-cpu.so - kai_rhs_pack_nxk_qsi4c32pscalef16_qsu4c32s16s0.c:107-154 [...] | kai_run_rhs_pack_nxk_qsi4c32pscalef16_qsu4c32s16s0 | InBetween | 5.80 | 0.14 | 0.14 | 0.26 | 0.26 | 0.00 | 0.00 | 1 | 0 | 44.08 | 1 | 1 | 2 | 1 | NA | NA | NA | NA | NA | 0.00 |
| 785 | libggml-cpu.so - vec.cpp:311-337 [...] | ggml_vec_dot_f16 | Outermost | 0.22 | 0.13 | 0.13 | 0.01 | 0.01 | 0.00 | 0.00 | 37 | 13.85 | 38.32 | 2.1 | 1.26 | 2.09 | 1.57 | NA | NA | NA | NA | NA | 0.00 |
| 815 | libggml-cpu.so - ops.cpp:6210-6475 [...] | ggml_compute_forward_rope_f32(ggml_compute_params const*, ggml_tensor*, bool) | InBetween | 0.33 | 0.12 | 1.90 | 0.01 | 0.06 | 0.00 | 0.05 | 32 | 0 | 26.47 | 2.04 | 1.52 | 4.42 | 2.13 | NA | NA | NA | NA | NA | 0.00 |
| 816 | libggml-cpu.so - ops.cpp:6210-6407 [...] | ggml_compute_forward_rope_f32(ggml_compute_params const*, ggml_tensor*, bool) | InBetween | 0.33 | 0.09 | 1.76 | 0.01 | 0.06 | 0.00 | 0.05 | 28 | 0 | 29.46 | 2.58 | 1.22 | 3.28 | 2.55 | NA | NA | NA | NA | NA | 0.00 |
| 830 | libggml-cpu.so - ops.cpp:6446-6457 [...] | ggml_compute_forward_rope_f32(ggml_compute_params const*, ggml_tensor*, bool) | Innermost | 0.22 | 0.08 | 0.08 | 0.01 | 0.01 | 0.00 | 0.00 | 29 | 42.86 | 53.57 | 1 | 1 | 2.06 | 1.87 | 1 | 0 | 0 | 15 | 0 | 53.13 |
| 1611 | libggml-cpu.so - ops.cpp:8778-8939 [...] | ggml_compute_forward_flash_attn_ext | Outermost | 0.22 | 0.07 | 1.55 | 0.01 | 0.05 | 0.00 | 0.04 | 23 | 0 | 28.85 | 3.22 | 1.34 | 4.47 | 1.92 | NA | NA | NA | NA | NA | 0.00 |
| 2220 | libggml-cpu.so - kai_rhs_pack_nxk_qsi4c32pscalef16_qsu4c32s16s0.c:107-154 [...] | kai_run_rhs_pack_nxk_qsi4c32pscalef16_qsu4c32s16s0 | InBetween | 2.30 | 0.06 | 0.36 | 0.10 | 0.66 | 0.00 | 0.01 | 1 | 0 | 41.07 | 1 | 1 | 3.03 | 1 | NA | NA | NA | NA | NA | 0.00 |
| 1197 | libggml-cpu.so - vec.h:677-677 [...] | ggml_compute_forward_rms_norm | Innermost | 0.11 | 0.05 | 0.05 | 0.01 | 0.01 | 0.00 | 0.00 | 19 | 44.44 | 100 | 1 | 1 | 1 | 1 | 0 | 0 | 1 | 3 | 0 | 56.25 |
| 1615 | libggml-cpu.so - ops.cpp:8778-8939 [...] | ggml_compute_forward_flash_attn_ext | InBetween | 0.22 | 0.05 | 0.08 | 0.01 | 0.02 | 0.00 | 0.00 | 17 | 0 | 41.2 | 2.48 | 1.18 | 1.88 | 1.89 | NA | NA | NA | NA | NA | 0.00 |
| 2114 | libggml-cpu.so - quants.c:2492-2654 [...] | ggml_vec_dot_q6_K_q8_K | InBetween | 0.11 | 0.04 | 0.24 | 0.01 | 0.01 | 0.00 | 0.01 | 16 | 24.64 | 46.2 | 1.04 | 1 | 1.77 | 1 | 1 | 3 | 0 | 5 | 0 | 72.22 |
| 1612 | libggml-cpu.so - ops.cpp:8818-8826 [...] | ggml_compute_forward_flash_attn_ext | Innermost | 0.22 | 0.03 | 0.03 | 0.01 | 0.01 | 0.00 | 0.00 | 10 | 0 | 41.07 | 10.67 | 1 | 5.33 | 1.67 | NA | NA | NA | NA | NA | 0.00 |
| 75 | libggml-cpu.so - ggml-cpu.c:1664-2898 [...] | ggml_graph_compute_thread | Outermost | 0.11 | 0.02 | 0.02 | 0.01 | 0.01 | 0.00 | 0.00 | 7 | 0 | 34.38 | 1 | 1 | 4.62 | 1 | NA | NA | NA | NA | NA | 0.00 |
| 1614 | libggml-cpu.so - ops.cpp:8778-8939 [...] | ggml_compute_forward_flash_attn_ext | InBetween | 0.11 | 0.02 | 0.03 | 0.01 | 0.01 | 0.00 | 0.00 | 7 | 18.93 | 59.1 | 4.46 | 1.02 | 1.08 | 1 | NA | NA | NA | NA | NA | 0.00 |
| 817 | libggml-cpu.so - ops.cpp:6210-6439 [...] | ggml_compute_forward_rope_f32(ggml_compute_params const*, ggml_tensor*, bool) | InBetween | 0.11 | 0.02 | 1.78 | 0.01 | 0.05 | 0.00 | 0.05 | 6 | 0 | 49.47 | 1 | 1 | 1.81 | 1 | NA | NA | NA | NA | NA | 0.00 |
| 821 | libggml-cpu.so - ops.cpp:6210-6490 [...] | ggml_compute_forward_rope_f32(ggml_compute_params const*, ggml_tensor*, bool) | Outermost | 0.11 | 0.01 | 2.00 | 0.01 | 0.05 | 0.00 | 0.06 | 5 | 13.46 | 34.38 | 1.56 | 1 | 1.6 | 1 | NA | NA | NA | NA | NA | 0.00 |
| 1617 | libggml-cpu.so - vec.h:677-677 [...] | ggml_compute_forward_flash_attn_ext | Innermost | 0.11 | 0.01 | 0.01 | 0.01 | 0.01 | 0.00 | 0.00 | 4 | 44.44 | 100 | 1 | 1 | 1 | 1 | 0 | 0 | 1 | 3 | 0 | 56.25 |
| 1191 | libggml-cpu.so - ops.cpp:4319-4351 [...] | ggml_compute_forward_rms_norm | InBetween | 0.11 | 0.01 | 0.49 | 0.00 | 0.02 | 0.00 | 0.01 | 4 | 24.28 | 65.39 | 1.71 | 1.04 | 1.28 | 1 | NA | NA | NA | NA | NA | 0.00 |