Help is available by moving the cursor above any 
symbol or by checking MAQAO website.
[ 4 / 4 ] Application profile is long enough (1007.68 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 0 / 3 ] Some functions are compiled with a low optimization level (O0 or O1)
To have better performances, it is advised to help the compiler by using a proper optimization level (-O2 of higher). Warning, depending on compilers, faster optimization levels can decrease numeric accuracy.
[ 0 / 3 ] Most of time spent in analyzed modules comes from functions without compilation information
Functions without compilation information (typically not compiled with -g and -grecord-gcc-switches) cumulate 100.00% of the time spent in analyzed modules. Check that -g and -grecord-gcc-switches are present. Remark: if -g and -grecord-gcc-switches are indeed used, this can also be due to some compiler built-in functions (typically math) or statically linked libraries. This warning can be ignored in that case.
[ 3 / 3 ] Host configuration allows retrieval of all necessary metrics.
[ 0 / 3 ] Compilation of some functions is not optimized for the target processor
Architecture specific options are needed to produce efficient code for a specific processor ( -x(target) or -ax(target) ).
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 0.06 % 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
[ 1 / 1 ] Lstopo present. The Topology lstopo report will be generated.
[ 0 / 0 ] Fastmath not used
Consider to add ffast-math to compilation flags (or replace -O3 with -Ofast) to unlock potential extra speedup by relaxing floating-point computation consistency. Warning: floating-point accuracy may be reduced and the compliance to IEEE/ISO rules/specifications for math functions will be relaxed, typically 'errno' will no longer be set after calling some math functions.
[ 4 / 4 ] Enough time of the experiment time spent in analyzed loops (98.31%)
If the time spent in analyzed loops is less than 30%, standard loop optimizations will have a limited impact on application performances.
[ 4 / 4 ] Threads activity is good
On average, more than 99.92% of observed threads are actually active
[ 4 / 4 ] CPU activity is good
CPU cores are active 99.93% of time
[ 4 / 4 ] Loop profile is not flat
At least one loop coverage is greater than 4% (59.53%), representing an hotspot for the application
[ 4 / 4 ] Enough time of the experiment time spent in analyzed innermost loops (37.35%)
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.
[ 4 / 4 ] Affinity is good (99.99%)
Threads are not migrating to CPU cores: probably successfully pinned
[ 3 / 3 ] Less than 10% (0.00%) is spend in BLAS1 operations
It could be more efficient to inline by hand BLAS1 operations
[ 3 / 3 ] Functions mostly use all threads
Functions running on a reduced number of threads (typically sequential code) cover less than 10% of application walltime (0.01%)
[ 0 / 3 ] Cumulative Outermost/In between loops coverage (60.96%) greater than cumulative innermost loop coverage (37.35%)
Having cumulative Outermost/In between loops coverage greater than cumulative innermost loop coverage will make loop optimization more complex
[ 2 / 2 ] Less than 10% (0.00%) is spend in BLAS2 operations
BLAS2 calls usually could make a poor cache usage and could benefit from inlining.
[ 2 / 2 ] Less than 10% (1.37%) is spend in Libm/SVML (special functions)
| Loop ID | Analysis | Penalty Score |
|---|---|---|
| ►Loop 20 - md-clang-O3-ffast-math | Execution Time: 59 % - Vectorization Ratio: 38.44 % - Vector Length Use: 17.31 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [SA] Presence of expensive FP instructions - Perform hoisting, change algorithm, use SVML or proper numerical library or perform value profiling (count the number of distinct input values). There are 1 issues (= instructions) costing 4 points each. | 4 |
| ○ | [SA] Presence of a large number of scalar integer instructions - Simplify loop structure, perform loop splitting or perform unroll and jam. This issue costs 2 points. | 2 |
| ►Control Flow Issues | 4 | |
| ○ | [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] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Data Access Issues | 22 | |
| ○ | [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 3 issues ( = data accesses) costing 2 point each. | 6 |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 3 issues ( = indirect data accesses) costing 4 point each. | 12 |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, SHUFFLE/PERM) - Simplify data access and try to get stride 1 access. There are 4 issues (= instructions) costing 1 point each. | 4 |
| ►Vectorization Roadblocks | 22 | |
| ○ | [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 3 issues ( = data accesses) costing 2 point each. | 6 |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 3 issues ( = indirect data accesses) costing 4 point each. | 12 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Inefficient Vectorization | 4 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, SHUFFLE/PERM) - Simplify data access and try to get stride 1 access. There are 4 issues (= instructions) costing 1 point each. | 4 |
| ►Loop 26 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 25 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 22 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 23 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 21 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 24 - md-clang-O3-ffast-math | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 19 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 11.67 % | |
| ►Control Flow Issues | 5 | |
| ○ | [SA] Several paths (3 paths) - Simplify control structure or force the compiler to use masked instructions. There are 3 issues ( = paths) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 5 | |
| ○ | [SA] Several paths (3 paths) - Simplify control structure or force the compiler to use masked instructions. There are 3 issues ( = paths) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 18 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 7.89 % | |
| ►Control Flow Issues | 12 | |
| ○ | [SA] Too many paths (6 paths) - Simplify control structure. There are 6 issues ( = paths) costing 1 point each with a malus of 4 points. | 10 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 12 | |
| ○ | [SA] Too many paths (6 paths) - Simplify control structure. There are 6 issues ( = paths) costing 1 point each with a malus of 4 points. | 10 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 36 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 75.00 % - Vector Length Use: 29.21 % | |
| ►Loop Computation Issues | 2 | |
| ○ | [SA] Presence of a large number of scalar integer instructions - Simplify loop structure, perform loop splitting or perform unroll and jam. This issue costs 2 points. | 2 |
| ►Control Flow Issues | 1 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 1 issues (= calls) costing 1 point each. | 1 |
| ►Data Access Issues | 22 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, BLEND/MERGE, SHUFFLE/PERM, BROADCAST) - Simplify data access and try to get stride 1 access. There are 20 issues (= instructions) costing 1 point each. | 20 |
| ○ | [SA] More than 20% of the loads are accessing the stack - Perform loop splitting to decrease pressure on registers. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 1 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 1 issues (= calls) costing 1 point each. | 1 |
| ►Inefficient Vectorization | 20 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, BLEND/MERGE, SHUFFLE/PERM, BROADCAST) - Simplify data access and try to get stride 1 access. There are 20 issues (= instructions) costing 1 point each. | 20 |
[ 4 / 4 ] Application profile is long enough (511.80 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 0 / 3 ] Some functions are compiled with a low optimization level (O0 or O1)
To have better performances, it is advised to help the compiler by using a proper optimization level (-O2 of higher). Warning, depending on compilers, faster optimization levels can decrease numeric accuracy.
[ 0 / 3 ] Most of time spent in analyzed modules comes from functions without compilation information
Functions without compilation information (typically not compiled with -g and -grecord-gcc-switches) cumulate 100.00% of the time spent in analyzed modules. Check that -g and -grecord-gcc-switches are present. Remark: if -g and -grecord-gcc-switches are indeed used, this can also be due to some compiler built-in functions (typically math) or statically linked libraries. This warning can be ignored in that case.
[ 3 / 3 ] Host configuration allows retrieval of all necessary metrics.
[ 0 / 3 ] Compilation of some functions is not optimized for the target processor
Architecture specific options are needed to produce efficient code for a specific processor ( -x(target) or -ax(target) ).
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 0.05 % 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
[ 1 / 1 ] Lstopo present. The Topology lstopo report will be generated.
[ 0 / 0 ] Fastmath not used
Consider to add ffast-math to compilation flags (or replace -O3 with -Ofast) to unlock potential extra speedup by relaxing floating-point computation consistency. Warning: floating-point accuracy may be reduced and the compliance to IEEE/ISO rules/specifications for math functions will be relaxed, typically 'errno' will no longer be set after calling some math functions.
[ 4 / 4 ] Enough time of the experiment time spent in analyzed loops (97.56%)
If the time spent in analyzed loops is less than 30%, standard loop optimizations will have a limited impact on application performances.
[ 4 / 4 ] Threads activity is good
On average, more than 199.83% of observed threads are actually active
[ 4 / 4 ] CPU activity is good
CPU cores are active 99.93% of time
[ 4 / 4 ] Loop profile is not flat
At least one loop coverage is greater than 4% (58.81%), representing an hotspot for the application
[ 4 / 4 ] Enough time of the experiment time spent in analyzed innermost loops (37.23%)
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.
[ 4 / 4 ] Affinity is good (99.99%)
Threads are not migrating to CPU cores: probably successfully pinned
[ 3 / 3 ] Less than 10% (0.00%) is spend in BLAS1 operations
It could be more efficient to inline by hand BLAS1 operations
[ 3 / 3 ] Functions mostly use all threads
Functions running on a reduced number of threads (typically sequential code) cover less than 10% of application walltime (0.20%)
[ 0 / 3 ] Cumulative Outermost/In between loops coverage (60.34%) greater than cumulative innermost loop coverage (37.23%)
Having cumulative Outermost/In between loops coverage greater than cumulative innermost loop coverage will make loop optimization more complex
[ 2 / 2 ] Less than 10% (0.00%) is spend in BLAS2 operations
BLAS2 calls usually could make a poor cache usage and could benefit from inlining.
[ 2 / 2 ] Less than 10% (1.34%) is spend in Libm/SVML (special functions)
| Loop ID | Analysis | Penalty Score |
|---|---|---|
| ►Loop 20 - md-clang-O3-ffast-math | Execution Time: 58 % - Vectorization Ratio: 38.44 % - Vector Length Use: 17.31 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [SA] Presence of expensive FP instructions - Perform hoisting, change algorithm, use SVML or proper numerical library or perform value profiling (count the number of distinct input values). There are 1 issues (= instructions) costing 4 points each. | 4 |
| ○ | [SA] Presence of a large number of scalar integer instructions - Simplify loop structure, perform loop splitting or perform unroll and jam. This issue costs 2 points. | 2 |
| ►Control Flow Issues | 4 | |
| ○ | [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] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Data Access Issues | 22 | |
| ○ | [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 3 issues ( = data accesses) costing 2 point each. | 6 |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 3 issues ( = indirect data accesses) costing 4 point each. | 12 |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, SHUFFLE/PERM) - Simplify data access and try to get stride 1 access. There are 4 issues (= instructions) costing 1 point each. | 4 |
| ►Vectorization Roadblocks | 22 | |
| ○ | [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 3 issues ( = data accesses) costing 2 point each. | 6 |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 3 issues ( = indirect data accesses) costing 4 point each. | 12 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Inefficient Vectorization | 4 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, SHUFFLE/PERM) - Simplify data access and try to get stride 1 access. There are 4 issues (= instructions) costing 1 point each. | 4 |
| ►Loop 26 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 25 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 22 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 23 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 21 - md-clang-O3-ffast-math | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 24 - md-clang-O3-ffast-math | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 19 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 11.67 % | |
| ►Control Flow Issues | 5 | |
| ○ | [SA] Several paths (3 paths) - Simplify control structure or force the compiler to use masked instructions. There are 3 issues ( = paths) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 5 | |
| ○ | [SA] Several paths (3 paths) - Simplify control structure or force the compiler to use masked instructions. There are 3 issues ( = paths) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 18 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 7.89 % | |
| ►Control Flow Issues | 12 | |
| ○ | [SA] Too many paths (6 paths) - Simplify control structure. There are 6 issues ( = paths) costing 1 point each with a malus of 4 points. | 10 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 12 | |
| ○ | [SA] Too many paths (6 paths) - Simplify control structure. There are 6 issues ( = paths) costing 1 point each with a malus of 4 points. | 10 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 36 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 75.00 % - Vector Length Use: 29.21 % | |
| ►Loop Computation Issues | 2 | |
| ○ | [SA] Presence of a large number of scalar integer instructions - Simplify loop structure, perform loop splitting or perform unroll and jam. This issue costs 2 points. | 2 |
| ►Control Flow Issues | 1 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 1 issues (= calls) costing 1 point each. | 1 |
| ►Data Access Issues | 22 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, BLEND/MERGE, SHUFFLE/PERM, BROADCAST) - Simplify data access and try to get stride 1 access. There are 20 issues (= instructions) costing 1 point each. | 20 |
| ○ | [SA] More than 20% of the loads are accessing the stack - Perform loop splitting to decrease pressure on registers. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 1 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 1 issues (= calls) costing 1 point each. | 1 |
| ►Inefficient Vectorization | 20 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, BLEND/MERGE, SHUFFLE/PERM, BROADCAST) - Simplify data access and try to get stride 1 access. There are 20 issues (= instructions) costing 1 point each. | 20 |
[ 4 / 4 ] Application profile is long enough (263.97 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 0 / 3 ] Some functions are compiled with a low optimization level (O0 or O1)
To have better performances, it is advised to help the compiler by using a proper optimization level (-O2 of higher). Warning, depending on compilers, faster optimization levels can decrease numeric accuracy.
[ 0 / 3 ] Most of time spent in analyzed modules comes from functions without compilation information
Functions without compilation information (typically not compiled with -g and -grecord-gcc-switches) cumulate 100.00% of the time spent in analyzed modules. Check that -g and -grecord-gcc-switches are present. Remark: if -g and -grecord-gcc-switches are indeed used, this can also be due to some compiler built-in functions (typically math) or statically linked libraries. This warning can be ignored in that case.
[ 3 / 3 ] Host configuration allows retrieval of all necessary metrics.
[ 0 / 3 ] Compilation of some functions is not optimized for the target processor
Architecture specific options are needed to produce efficient code for a specific processor ( -x(target) or -ax(target) ).
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 0.05 % 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
[ 1 / 1 ] Lstopo present. The Topology lstopo report will be generated.
[ 0 / 0 ] Fastmath not used
Consider to add ffast-math to compilation flags (or replace -O3 with -Ofast) to unlock potential extra speedup by relaxing floating-point computation consistency. Warning: floating-point accuracy may be reduced and the compliance to IEEE/ISO rules/specifications for math functions will be relaxed, typically 'errno' will no longer be set after calling some math functions.
[ 4 / 4 ] Enough time of the experiment time spent in analyzed loops (95.36%)
If the time spent in analyzed loops is less than 30%, standard loop optimizations will have a limited impact on application performances.
[ 4 / 4 ] Threads activity is good
On average, more than 399.40% of observed threads are actually active
[ 4 / 4 ] CPU activity is good
CPU cores are active 99.88% of time
[ 4 / 4 ] Loop profile is not flat
At least one loop coverage is greater than 4% (57.12%), representing an hotspot for the application
[ 4 / 4 ] Enough time of the experiment time spent in analyzed innermost loops (36.75%)
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.
[ 4 / 4 ] Affinity is good (99.99%)
Threads are not migrating to CPU cores: probably successfully pinned
[ 3 / 3 ] Less than 10% (0.00%) is spend in BLAS1 operations
It could be more efficient to inline by hand BLAS1 operations
[ 3 / 3 ] Functions mostly use all threads
Functions running on a reduced number of threads (typically sequential code) cover less than 10% of application walltime (0.44%)
[ 0 / 3 ] Cumulative Outermost/In between loops coverage (58.61%) greater than cumulative innermost loop coverage (36.75%)
Having cumulative Outermost/In between loops coverage greater than cumulative innermost loop coverage will make loop optimization more complex
[ 2 / 2 ] Less than 10% (0.00%) is spend in BLAS2 operations
BLAS2 calls usually could make a poor cache usage and could benefit from inlining.
[ 2 / 2 ] Less than 10% (1.31%) is spend in Libm/SVML (special functions)
| Loop ID | Analysis | Penalty Score |
|---|---|---|
| ►Loop 20 - md-clang-O3-ffast-math | Execution Time: 57 % - Vectorization Ratio: 38.44 % - Vector Length Use: 17.31 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [SA] Presence of expensive FP instructions - Perform hoisting, change algorithm, use SVML or proper numerical library or perform value profiling (count the number of distinct input values). There are 1 issues (= instructions) costing 4 points each. | 4 |
| ○ | [SA] Presence of a large number of scalar integer instructions - Simplify loop structure, perform loop splitting or perform unroll and jam. This issue costs 2 points. | 2 |
| ►Control Flow Issues | 4 | |
| ○ | [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] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Data Access Issues | 22 | |
| ○ | [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 3 issues ( = data accesses) costing 2 point each. | 6 |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 3 issues ( = indirect data accesses) costing 4 point each. | 12 |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, SHUFFLE/PERM) - Simplify data access and try to get stride 1 access. There are 4 issues (= instructions) costing 1 point each. | 4 |
| ►Vectorization Roadblocks | 22 | |
| ○ | [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 3 issues ( = data accesses) costing 2 point each. | 6 |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 3 issues ( = indirect data accesses) costing 4 point each. | 12 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Inefficient Vectorization | 4 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, SHUFFLE/PERM) - Simplify data access and try to get stride 1 access. There are 4 issues (= instructions) costing 1 point each. | 4 |
| ►Loop 26 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 25 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 22 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 23 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 21 - md-clang-O3-ffast-math | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 24 - md-clang-O3-ffast-math | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 19 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 11.67 % | |
| ►Control Flow Issues | 5 | |
| ○ | [SA] Several paths (3 paths) - Simplify control structure or force the compiler to use masked instructions. There are 3 issues ( = paths) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 5 | |
| ○ | [SA] Several paths (3 paths) - Simplify control structure or force the compiler to use masked instructions. There are 3 issues ( = paths) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 18 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 7.89 % | |
| ►Control Flow Issues | 12 | |
| ○ | [SA] Too many paths (6 paths) - Simplify control structure. There are 6 issues ( = paths) costing 1 point each with a malus of 4 points. | 10 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 12 | |
| ○ | [SA] Too many paths (6 paths) - Simplify control structure. There are 6 issues ( = paths) costing 1 point each with a malus of 4 points. | 10 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 36 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 75.00 % - Vector Length Use: 29.21 % | |
| ►Loop Computation Issues | 2 | |
| ○ | [SA] Presence of a large number of scalar integer instructions - Simplify loop structure, perform loop splitting or perform unroll and jam. This issue costs 2 points. | 2 |
| ►Control Flow Issues | 1 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 1 issues (= calls) costing 1 point each. | 1 |
| ►Data Access Issues | 22 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, BLEND/MERGE, SHUFFLE/PERM, BROADCAST) - Simplify data access and try to get stride 1 access. There are 20 issues (= instructions) costing 1 point each. | 20 |
| ○ | [SA] More than 20% of the loads are accessing the stack - Perform loop splitting to decrease pressure on registers. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 1 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 1 issues (= calls) costing 1 point each. | 1 |
| ►Inefficient Vectorization | 20 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, BLEND/MERGE, SHUFFLE/PERM, BROADCAST) - Simplify data access and try to get stride 1 access. There are 20 issues (= instructions) costing 1 point each. | 20 |
[ 4 / 4 ] Application profile is long enough (137.85 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 0 / 3 ] Some functions are compiled with a low optimization level (O0 or O1)
To have better performances, it is advised to help the compiler by using a proper optimization level (-O2 of higher). Warning, depending on compilers, faster optimization levels can decrease numeric accuracy.
[ 0 / 3 ] Most of time spent in analyzed modules comes from functions without compilation information
Functions without compilation information (typically not compiled with -g and -grecord-gcc-switches) cumulate 100.00% of the time spent in analyzed modules. Check that -g and -grecord-gcc-switches are present. Remark: if -g and -grecord-gcc-switches are indeed used, this can also be due to some compiler built-in functions (typically math) or statically linked libraries. This warning can be ignored in that case.
[ 3 / 3 ] Host configuration allows retrieval of all necessary metrics.
[ 0 / 3 ] Compilation of some functions is not optimized for the target processor
Architecture specific options are needed to produce efficient code for a specific processor ( -x(target) or -ax(target) ).
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 0.05 % 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
[ 1 / 1 ] Lstopo present. The Topology lstopo report will be generated.
[ 0 / 0 ] Fastmath not used
Consider to add ffast-math to compilation flags (or replace -O3 with -Ofast) to unlock potential extra speedup by relaxing floating-point computation consistency. Warning: floating-point accuracy may be reduced and the compliance to IEEE/ISO rules/specifications for math functions will be relaxed, typically 'errno' will no longer be set after calling some math functions.
[ 4 / 4 ] Enough time of the experiment time spent in analyzed loops (92.57%)
If the time spent in analyzed loops is less than 30%, standard loop optimizations will have a limited impact on application performances.
[ 4 / 4 ] Threads activity is good
On average, more than 797.62% of observed threads are actually active
[ 4 / 4 ] CPU activity is good
CPU cores are active 99.77% of time
[ 4 / 4 ] Loop profile is not flat
At least one loop coverage is greater than 4% (54.80%), representing an hotspot for the application
[ 4 / 4 ] Enough time of the experiment time spent in analyzed innermost loops (36.39%)
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.
[ 4 / 4 ] Affinity is good (99.99%)
Threads are not migrating to CPU cores: probably successfully pinned
[ 3 / 3 ] Less than 10% (0.00%) is spend in BLAS1 operations
It could be more efficient to inline by hand BLAS1 operations
[ 3 / 3 ] Functions mostly use all threads
Functions running on a reduced number of threads (typically sequential code) cover less than 10% of application walltime (0.88%)
[ 0 / 3 ] Cumulative Outermost/In between loops coverage (56.19%) greater than cumulative innermost loop coverage (36.39%)
Having cumulative Outermost/In between loops coverage greater than cumulative innermost loop coverage will make loop optimization more complex
[ 2 / 2 ] Less than 10% (0.00%) is spend in BLAS2 operations
BLAS2 calls usually could make a poor cache usage and could benefit from inlining.
[ 2 / 2 ] Less than 10% (1.25%) is spend in Libm/SVML (special functions)
| Loop ID | Analysis | Penalty Score |
|---|---|---|
| ►Loop 20 - md-clang-O3-ffast-math | Execution Time: 54 % - Vectorization Ratio: 38.44 % - Vector Length Use: 17.31 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [SA] Presence of expensive FP instructions - Perform hoisting, change algorithm, use SVML or proper numerical library or perform value profiling (count the number of distinct input values). There are 1 issues (= instructions) costing 4 points each. | 4 |
| ○ | [SA] Presence of a large number of scalar integer instructions - Simplify loop structure, perform loop splitting or perform unroll and jam. This issue costs 2 points. | 2 |
| ►Control Flow Issues | 4 | |
| ○ | [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] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Data Access Issues | 22 | |
| ○ | [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 3 issues ( = data accesses) costing 2 point each. | 6 |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 3 issues ( = indirect data accesses) costing 4 point each. | 12 |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, SHUFFLE/PERM) - Simplify data access and try to get stride 1 access. There are 4 issues (= instructions) costing 1 point each. | 4 |
| ►Vectorization Roadblocks | 22 | |
| ○ | [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 3 issues ( = data accesses) costing 2 point each. | 6 |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 3 issues ( = indirect data accesses) costing 4 point each. | 12 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Inefficient Vectorization | 4 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, SHUFFLE/PERM) - Simplify data access and try to get stride 1 access. There are 4 issues (= instructions) costing 1 point each. | 4 |
| ►Loop 26 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 25 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 22 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 23 - md-clang-O3-ffast-math | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 21 - md-clang-O3-ffast-math | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 24 - md-clang-O3-ffast-math | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 19 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 11.67 % | |
| ►Control Flow Issues | 5 | |
| ○ | [SA] Several paths (3 paths) - Simplify control structure or force the compiler to use masked instructions. There are 3 issues ( = paths) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 5 | |
| ○ | [SA] Several paths (3 paths) - Simplify control structure or force the compiler to use masked instructions. There are 3 issues ( = paths) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 18 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 7.89 % | |
| ►Control Flow Issues | 12 | |
| ○ | [SA] Too many paths (6 paths) - Simplify control structure. There are 6 issues ( = paths) costing 1 point each with a malus of 4 points. | 10 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 12 | |
| ○ | [SA] Too many paths (6 paths) - Simplify control structure. There are 6 issues ( = paths) costing 1 point each with a malus of 4 points. | 10 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 36 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 75.00 % - Vector Length Use: 29.21 % | |
| ►Loop Computation Issues | 2 | |
| ○ | [SA] Presence of a large number of scalar integer instructions - Simplify loop structure, perform loop splitting or perform unroll and jam. This issue costs 2 points. | 2 |
| ►Control Flow Issues | 1 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 1 issues (= calls) costing 1 point each. | 1 |
| ►Data Access Issues | 22 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, BLEND/MERGE, SHUFFLE/PERM, BROADCAST) - Simplify data access and try to get stride 1 access. There are 20 issues (= instructions) costing 1 point each. | 20 |
| ○ | [SA] More than 20% of the loads are accessing the stack - Perform loop splitting to decrease pressure on registers. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 1 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 1 issues (= calls) costing 1 point each. | 1 |
| ►Inefficient Vectorization | 20 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, BLEND/MERGE, SHUFFLE/PERM, BROADCAST) - Simplify data access and try to get stride 1 access. There are 20 issues (= instructions) costing 1 point each. | 20 |
[ 4 / 4 ] Application profile is long enough (89.53 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 0 / 3 ] Some functions are compiled with a low optimization level (O0 or O1)
To have better performances, it is advised to help the compiler by using a proper optimization level (-O2 of higher). Warning, depending on compilers, faster optimization levels can decrease numeric accuracy.
[ 0 / 3 ] Most of time spent in analyzed modules comes from functions without compilation information
Functions without compilation information (typically not compiled with -g and -grecord-gcc-switches) cumulate 100.00% of the time spent in analyzed modules. Check that -g and -grecord-gcc-switches are present. Remark: if -g and -grecord-gcc-switches are indeed used, this can also be due to some compiler built-in functions (typically math) or statically linked libraries. This warning can be ignored in that case.
[ 3 / 3 ] Host configuration allows retrieval of all necessary metrics.
[ 0 / 3 ] Compilation of some functions is not optimized for the target processor
Architecture specific options are needed to produce efficient code for a specific processor ( -x(target) or -ax(target) ).
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 0.05 % 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
[ 1 / 1 ] Lstopo present. The Topology lstopo report will be generated.
[ 0 / 0 ] Fastmath not used
Consider to add ffast-math to compilation flags (or replace -O3 with -Ofast) to unlock potential extra speedup by relaxing floating-point computation consistency. Warning: floating-point accuracy may be reduced and the compliance to IEEE/ISO rules/specifications for math functions will be relaxed, typically 'errno' will no longer be set after calling some math functions.
[ 4 / 4 ] Enough time of the experiment time spent in analyzed loops (89.06%)
If the time spent in analyzed loops is less than 30%, standard loop optimizations will have a limited impact on application performances.
[ 4 / 4 ] Threads activity is good
On average, more than 1293.43% of observed threads are actually active
[ 4 / 4 ] CPU activity is good
CPU cores are active 99.60% of time
[ 4 / 4 ] Loop profile is not flat
At least one loop coverage is greater than 4% (51.59%), representing an hotspot for the application
[ 4 / 4 ] Enough time of the experiment time spent in analyzed innermost loops (36.14%)
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.
[ 4 / 4 ] Affinity is good (99.98%)
Threads are not migrating to CPU cores: probably successfully pinned
[ 3 / 3 ] Less than 10% (0.00%) is spend in BLAS1 operations
It could be more efficient to inline by hand BLAS1 operations
[ 3 / 3 ] Functions mostly use all threads
Functions running on a reduced number of threads (typically sequential code) cover less than 10% of application walltime (1.57%)
[ 0 / 3 ] Cumulative Outermost/In between loops coverage (52.92%) greater than cumulative innermost loop coverage (36.14%)
Having cumulative Outermost/In between loops coverage greater than cumulative innermost loop coverage will make loop optimization more complex
[ 2 / 2 ] Less than 10% (0.00%) is spend in BLAS2 operations
BLAS2 calls usually could make a poor cache usage and could benefit from inlining.
[ 2 / 2 ] Less than 10% (1.22%) is spend in Libm/SVML (special functions)
| Loop ID | Analysis | Penalty Score |
|---|---|---|
| ►Loop 20 - md-clang-O3-ffast-math | Execution Time: 51 % - Vectorization Ratio: 38.44 % - Vector Length Use: 17.31 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [SA] Presence of expensive FP instructions - Perform hoisting, change algorithm, use SVML or proper numerical library or perform value profiling (count the number of distinct input values). There are 1 issues (= instructions) costing 4 points each. | 4 |
| ○ | [SA] Presence of a large number of scalar integer instructions - Simplify loop structure, perform loop splitting or perform unroll and jam. This issue costs 2 points. | 2 |
| ►Control Flow Issues | 4 | |
| ○ | [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] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Data Access Issues | 22 | |
| ○ | [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 3 issues ( = data accesses) costing 2 point each. | 6 |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 3 issues ( = indirect data accesses) costing 4 point each. | 12 |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, SHUFFLE/PERM) - Simplify data access and try to get stride 1 access. There are 4 issues (= instructions) costing 1 point each. | 4 |
| ►Vectorization Roadblocks | 22 | |
| ○ | [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 3 issues ( = data accesses) costing 2 point each. | 6 |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 3 issues ( = indirect data accesses) costing 4 point each. | 12 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Inefficient Vectorization | 4 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, SHUFFLE/PERM) - Simplify data access and try to get stride 1 access. There are 4 issues (= instructions) costing 1 point each. | 4 |
| ►Loop 26 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 25 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 22 - md-clang-O3-ffast-math | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 23 - md-clang-O3-ffast-math | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 21 - md-clang-O3-ffast-math | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 24 - md-clang-O3-ffast-math | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 19 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 11.67 % | |
| ►Control Flow Issues | 5 | |
| ○ | [SA] Several paths (3 paths) - Simplify control structure or force the compiler to use masked instructions. There are 3 issues ( = paths) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 5 | |
| ○ | [SA] Several paths (3 paths) - Simplify control structure or force the compiler to use masked instructions. There are 3 issues ( = paths) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 18 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 7.89 % | |
| ►Control Flow Issues | 12 | |
| ○ | [SA] Too many paths (6 paths) - Simplify control structure. There are 6 issues ( = paths) costing 1 point each with a malus of 4 points. | 10 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 12 | |
| ○ | [SA] Too many paths (6 paths) - Simplify control structure. There are 6 issues ( = paths) costing 1 point each with a malus of 4 points. | 10 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 36 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 75.00 % - Vector Length Use: 29.21 % | |
| ►Loop Computation Issues | 2 | |
| ○ | [SA] Presence of a large number of scalar integer instructions - Simplify loop structure, perform loop splitting or perform unroll and jam. This issue costs 2 points. | 2 |
| ►Control Flow Issues | 1 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 1 issues (= calls) costing 1 point each. | 1 |
| ►Data Access Issues | 22 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, BLEND/MERGE, SHUFFLE/PERM, BROADCAST) - Simplify data access and try to get stride 1 access. There are 20 issues (= instructions) costing 1 point each. | 20 |
| ○ | [SA] More than 20% of the loads are accessing the stack - Perform loop splitting to decrease pressure on registers. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 1 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 1 issues (= calls) costing 1 point each. | 1 |
| ►Inefficient Vectorization | 20 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, BLEND/MERGE, SHUFFLE/PERM, BROADCAST) - Simplify data access and try to get stride 1 access. There are 20 issues (= instructions) costing 1 point each. | 20 |
[ 4 / 4 ] Application profile is long enough (51.76 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 0 / 3 ] Some functions are compiled with a low optimization level (O0 or O1)
To have better performances, it is advised to help the compiler by using a proper optimization level (-O2 of higher). Warning, depending on compilers, faster optimization levels can decrease numeric accuracy.
[ 0 / 3 ] Most of time spent in analyzed modules comes from functions without compilation information
Functions without compilation information (typically not compiled with -g and -grecord-gcc-switches) cumulate 100.00% of the time spent in analyzed modules. Check that -g and -grecord-gcc-switches are present. Remark: if -g and -grecord-gcc-switches are indeed used, this can also be due to some compiler built-in functions (typically math) or statically linked libraries. This warning can be ignored in that case.
[ 3 / 3 ] Host configuration allows retrieval of all necessary metrics.
[ 0 / 3 ] Compilation of some functions is not optimized for the target processor
Architecture specific options are needed to produce efficient code for a specific processor ( -x(target) or -ax(target) ).
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 0.08 % 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
[ 1 / 1 ] Lstopo present. The Topology lstopo report will be generated.
[ 0 / 0 ] Fastmath not used
Consider to add ffast-math to compilation flags (or replace -O3 with -Ofast) to unlock potential extra speedup by relaxing floating-point computation consistency. Warning: floating-point accuracy may be reduced and the compliance to IEEE/ISO rules/specifications for math functions will be relaxed, typically 'errno' will no longer be set after calling some math functions.
[ 4 / 4 ] Enough time of the experiment time spent in analyzed loops (82.26%)
If the time spent in analyzed loops is less than 30%, standard loop optimizations will have a limited impact on application performances.
[ 4 / 4 ] Threads activity is good
On average, more than 2576.34% of observed threads are actually active
[ 4 / 4 ] CPU activity is good
CPU cores are active 99.29% of time
[ 4 / 4 ] Loop profile is not flat
At least one loop coverage is greater than 4% (46.78%), representing an hotspot for the application
[ 4 / 4 ] Enough time of the experiment time spent in analyzed innermost loops (34.42%)
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.
[ 4 / 4 ] Affinity is good (99.97%)
Threads are not migrating to CPU cores: probably successfully pinned
[ 3 / 3 ] Less than 10% (0.00%) is spend in BLAS1 operations
It could be more efficient to inline by hand BLAS1 operations
[ 3 / 3 ] Functions mostly use all threads
Functions running on a reduced number of threads (typically sequential code) cover less than 10% of application walltime (4.52%)
[ 0 / 3 ] Cumulative Outermost/In between loops coverage (47.84%) greater than cumulative innermost loop coverage (34.42%)
Having cumulative Outermost/In between loops coverage greater than cumulative innermost loop coverage will make loop optimization more complex
[ 2 / 2 ] Less than 10% (0.00%) is spend in BLAS2 operations
BLAS2 calls usually could make a poor cache usage and could benefit from inlining.
[ 2 / 2 ] Less than 10% (1.03%) is spend in Libm/SVML (special functions)
| Loop ID | Analysis | Penalty Score |
|---|---|---|
| ►Loop 20 - md-clang-O3-ffast-math | Execution Time: 46 % - Vectorization Ratio: 38.44 % - Vector Length Use: 17.31 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [SA] Presence of expensive FP instructions - Perform hoisting, change algorithm, use SVML or proper numerical library or perform value profiling (count the number of distinct input values). There are 1 issues (= instructions) costing 4 points each. | 4 |
| ○ | [SA] Presence of a large number of scalar integer instructions - Simplify loop structure, perform loop splitting or perform unroll and jam. This issue costs 2 points. | 2 |
| ►Control Flow Issues | 4 | |
| ○ | [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] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Data Access Issues | 22 | |
| ○ | [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 3 issues ( = data accesses) costing 2 point each. | 6 |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 3 issues ( = indirect data accesses) costing 4 point each. | 12 |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, SHUFFLE/PERM) - Simplify data access and try to get stride 1 access. There are 4 issues (= instructions) costing 1 point each. | 4 |
| ►Vectorization Roadblocks | 22 | |
| ○ | [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 3 issues ( = data accesses) costing 2 point each. | 6 |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 3 issues ( = indirect data accesses) costing 4 point each. | 12 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Inefficient Vectorization | 4 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT, SHUFFLE/PERM) - Simplify data access and try to get stride 1 access. There are 4 issues (= instructions) costing 1 point each. | 4 |
| ►Loop 25 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 26 - md-clang-O3-ffast-math | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 24 - md-clang-O3-ffast-math | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 22 - md-clang-O3-ffast-math | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 21 - md-clang-O3-ffast-math | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 23 - md-clang-O3-ffast-math | Execution Time: 4 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 4 | |
| ○ | [SA] Less than 10% of the FP ADD/SUB/MUL arithmetic operations are performed using FMA - Reorganize arithmetic expressions to exhibit potential for FMA. This issue costs 4 points. | 4 |
| ►Loop 19 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 11.67 % | |
| ►Control Flow Issues | 5 | |
| ○ | [SA] Several paths (3 paths) - Simplify control structure or force the compiler to use masked instructions. There are 3 issues ( = paths) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 5 | |
| ○ | [SA] Several paths (3 paths) - Simplify control structure or force the compiler to use masked instructions. There are 3 issues ( = paths) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 18 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 7.89 % | |
| ►Control Flow Issues | 12 | |
| ○ | [SA] Too many paths (6 paths) - Simplify control structure. There are 6 issues ( = paths) costing 1 point each with a malus of 4 points. | 10 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Vectorization Roadblocks | 12 | |
| ○ | [SA] Too many paths (6 paths) - Simplify control structure. There are 6 issues ( = paths) costing 1 point each with a malus of 4 points. | 10 |
| ○ | [SA] Non innermost loop (InBetween) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ○Loop 38 - md-clang-O3-ffast-math | Execution Time: 0 % - Vectorization Ratio: 100.00 % - Vector Length Use: 50.00 % |