Help is available by moving the cursor above any symbol or by checking MAQAO website.
[ 4 / 4 ] Application profile is long enough (97.54 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 3 / 3 ] Optimization level option -O2 is used
To have better performances, it is advised to help the compiler by using a proper optimization level (-O2)
[ 3 / 3 ] Helper debug compilation options -g and -fno-omit-frame-pointer are used
-g option gives access to debugging informations, such are source locations and -fno-omit-frame-pointer improve the accuracy of callchains found during the application profiling.
[ 3 / 3 ] Architecture specific option -march=skylake-avx512 is used
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 1.54 % of the execution time)
To have a representative profiling, it is advised that the category "Others" represents less than 20% of the execution time in order to analyze as much as possible of the user code
[ 4 / 4 ] Enough time of the experiment time spent in analyzed loops (76.82%)
If the time spent in analyzed loops is less than 30%, standard loop optimizations will have a limited impact on application performances.
[ 4 / 4 ] Loop profile is not flat
At least one loop coverage is greater than 4% (40.42%), representing an hotspot for the application
[ 4 / 4 ] Enough time of the experiment time spent in analyzed innermost loops (76.72%)
If the time spent in analyzed innermost loops is less than 15%, standard innermost loop optimizations such as vectorisation will have a limited impact on application performances.
[ 3 / 3 ] Less than 10% (0%) is spend in BLAS1 operations
It could be more efficient to inline by hand BLAS1 operations
[ 3 / 3 ] Cumulative Outermost/In between loops coverage (0.10%) lower than cumulative innermost loop coverage (76.72%)
Having cumulative Outermost/In between loops coverage greater than cumulative innermost loop coverage will make loop optimization more complex
[ 2 / 2 ] Less than 10% (0.02%) is spend in Libm/SVML (special functions)
[ 2 / 2 ] Less than 10% (1.52%) is spend in BLAS2 operations
BLAS2 calls usually could make a poor cache usage and could benefit from inlining.
Loop ID | Module | Analysis | Penalty Score | Coverage (%) | Vectorization Ratio (%) | Vector Length Use (%) |
---|---|---|---|---|---|---|
►1107 | exec | Partial or unexisting vectorization - Use pragma to force vectorization and check potential dependencies between array access. | 4 | 40.42 | 0.9 | 12.33 |
○ | [SA] Several paths (2 paths) - Simplify control structure or force the compiler to use masked instructions. There are 2 issues ( = paths) costing 1 point each. | 2 | ||||
○ | [SA] Presence of constant non unit stride data access - Use array restructuring, perform loop interchange or use gather instructions to lower a bit the cost. There are 1 issues ( = data accesses) costing 2 point each. | 2 | ||||
►661 | exec | Inefficient vectorization. | 17 | 15.87 | 100 | 100 |
○ | [DA] Ratio time (ORIG)/time (DL1) is greater than 3 (12.27) - Perform blocking. Perform array restructuring. There are 2 issues (= non unit stride or indirect memory access) costing 2 point each, with an additional malus of 13 points due to the ORIG/DL1 ratio. | 17 | ||||
►739 | exec | Inefficient vectorization. | 7 | 8.67 | 100 | 100 |
○ | [DA] Ratio time (ORIG)/time (DL1) is greater than 3 (4.62) - Perform blocking. Perform array restructuring. There are 1 issues (= non unit stride or indirect memory access) costing 2 point each, with an additional malus of 5 points due to the ORIG/DL1 ratio. | 7 | ||||
►1106 | exec | Partial or unexisting vectorization - Use pragma to force vectorization and check potential dependencies between array access. | 4 | 3.19 | 0.9 | 12.33 |
○ | [SA] Several paths (2 paths) - Simplify control structure or force the compiler to use masked instructions. There are 2 issues ( = paths) costing 1 point each. | 2 | ||||
○ | [SA] Presence of constant non unit stride data access - Use array restructuring, perform loop interchange or use gather instructions to lower a bit the cost. There are 1 issues ( = data accesses) costing 2 point each. | 2 | ||||
►597 | exec | Partial or unexisting vectorization - Use pragma to force vectorization and check potential dependencies between array access. | 4 | 1.23 | 0 | 12.5 |
○ | [SA] Presence of constant non unit stride data access - Use array restructuring, perform loop interchange or use gather instructions to lower a bit the cost. There are 2 issues ( = data accesses) costing 2 point each. | 4 | ||||
○838 | exec | Partial or unexisting vectorization - No issue detected | 0 | 0.76 | 0 | 12.5 |
►741 | exec | Partial or unexisting vectorization - Use pragma to force vectorization and check potential dependencies between array access. | 10 | 0.73 | 11.11 | 13.89 |
○ | [SA] Presence of constant non unit stride data access - Use array restructuring, perform loop interchange or use gather instructions to lower a bit the cost. There are 5 issues ( = data accesses) costing 2 point each. | 10 | ||||
►307 | exec | Partial or unexisting vectorization - Use pragma to force vectorization and check potential dependencies between array access. | 1000 | 0.72 | 0 | 10 |
○ | [SA] Too many paths (6561 paths) - Simplify control structure. There are 6561 issues ( = paths) costing 1 point, limited to 1000. | 1000 | ||||
○ | Warning! Some static analysis are missing because the loop has too many paths. Use a higher value for --maximal_path_number option. | 0 | ||||
►594 | exec | Partial or unexisting vectorization - Use pragma to force vectorization and check potential dependencies between array access. | 4 | 0.6 | 0 | 12.5 |
○ | [SA] Presence of constant non unit stride data access - Use array restructuring, perform loop interchange or use gather instructions to lower a bit the cost. There are 2 issues ( = data accesses) costing 2 point each. | 4 | ||||
►837 | exec | Partial or unexisting vectorization - Use pragma to force vectorization and check potential dependencies between array access. | 2 | 0.43 | 28.57 | 16.07 |
○ | [SA] Presence of constant non unit stride data access - Use array restructuring, perform loop interchange or use gather instructions to lower a bit the cost. There are 1 issues ( = data accesses) costing 2 point each. | 2 |