Help is available by moving the cursor above any 
symbol or by checking MAQAO website.
| Total Time (s) | 667.74 | |
| Profiled Time (s) | 666.22 | |
| Time in analyzed loops (%) | 61.3 | |
| Time in analyzed innermost loops (%) | 42.4 | |
| Time in user code (%) | 81.0 | |
| Compilation Options Score (%) | 50.0 | |
| Array Access Efficiency (%) | 88.7 | |
|
| Potential Speedups |
| Perfect Flow Complexity | 1.09 | |
| Perfect OpenMP + MPI + Pthread | 1.06 | |
| Perfect OpenMP + MPI + Pthread + Perfect Load Distribution | 1.26 | |
| No Scalar Integer | Potential Speedup | 1.04 | |
| Nb Loops to get 80% | 10 | |
| FP Vectorised | Potential Speedup | 1.68 | |
| Nb Loops to get 80% | 10 | |
| Fully Vectorised | Potential Speedup | 2.36 | |
| Nb Loops to get 80% | 12 | |
| FP Arithmetic Only | Potential Speedup | 1.11 | |
| Nb Loops to get 80% | 13 | |
| Source Object | Issue |
|---|
| ▼picongpu– | |
| ▼Vector.hpp– | |
| ○ | -march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native. |
| ○ | -funroll-loops is missing. |
| ▼random.tcc– | |
| ○ | -march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native. |
| ○ | -funroll-loops is missing. |
| ▼TaskKernelCpuOmp2Blocks.hpp– | |
| ○ | -march=x86-64 is used but it should be replaced by a more architecture specific option or -march=native. |
| ○ | -funroll-loops is missing. |
| Application | ./bin/picongpu | | |
| Timestamp | 2023-12-07 12:07:29 |
Universal Timestamp | 1701947249 |
| Number of processes observed | 1 |
Number of threads observed | 53 |
| Experiment Type | OpenMP; | | |
| Machine | skylake | | |
| Model Name | Intel(R) Xeon(R) Platinum 8170 CPU @ 2.10GHz | | |
| Architecture | x86_64 |
Micro Architecture | SKYLAKE |
| Cache Size | 36608 KB |
Number of Cores | 26 |
| OS Version | Linux 6.5.7-arch1-1 #1 SMP PREEMPT_DYNAMIC Tue, 10 Oct 2023 21:10:21 +0000 | | |
| Architecture used during static analysis | x86_64 |
Micro Architecture used during static analysis | SKYLAKE |
| Frequency Driver | intel_cpufreq |
Frequency Governor | performance |
| Huge Pages | always |
Hyperthreading | off |
| Number of sockets | 2 |
Number of cores per socket | 26 |
| Compilation Options | picongpu: GNU C++17 13.2.1 20230801 -mtune=generic -march=x86-64 -g -O3 -std=c++17 -fno-omit-frame-pointer -fopenmp | | |
| Dataset | |
| Run Command | <executable> -d 1 1 1 -g 128 128 32 --periodic 1 1 1 -s 1500 --fields_energy.period 20 |
| Number Processes | 1 |
| Number Nodes | 1 |
| Filter | Not Used |
| Profile Start | Not Used |
