| | | | | | | requested parallelism | walltime sum (s) | nb instances | any sync average per thread time (s) | any wait average per thread time (s) | parallelism overhead (%) | local speedup if perfectly balanced | global speedup if perfectly balanced |
|---|
| start addr | function name | source location | level | ancestor thread num | invoker | parallel or teams | orig_0 | orig_0 | orig_0 | orig_0 | orig_0 | orig_0 | orig_0 | orig_0 |
|---|
| libggml-cpu.so:0x142b0 | ggml_graph_compute | ggml-cpu.c:682 | 0 | 0 | runtime | parallel | 96 | 13.839 | 513 | 7.330 | 7.269 | 53.0 | 2.126 | 2.009 |
| libggml-cpu.so:0x44936 | ggml_backend_amx_convert_weight(ggml_tensor*, void const*, u... | mmq.cpp:2337 | 0 | 0 | runtime | parallel | 96 | 0.227 | 225 | 0.100 | 0.100 | 44.1 | 1.788 | 1.007 |
