Help is available by moving the cursor above any 
symbol or by checking MAQAO website.
[ 4 / 4 ] Application profile is long enough (411.95 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with source/debug info
-g option gives access to debugging informations, such are source locations.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with compilation options informations and -fno-omit-frame-pointer is present
-fno-omit-frame-pointer improves the accuracy of callchains found during the application profiling.
[ 3 / 3 ] Host configuration allows retrieval of all necessary metrics.
[ 0 / 3 ] Compilation of some functions is not optimized for the target processor
Application run on the GRANITE_RAPIDS micro-architecture while the code was specialized for sapphirerapids.
[ 3 / 3 ] Optimization level option is correctly used
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 0.00 % 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.
[ 0 / 4 ] Too little time of the experiment time spent in analyzed loops (20.08%)
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 96.35% of observed threads are actually active
[ 4 / 4 ] CPU activity is good
CPU cores are active 96.35% of time
[ 4 / 4 ] Loop profile is not flat
At least one loop coverage is greater than 4% (7.80%), representing an hotspot for the application
[ 0 / 4 ] Too little time of the experiment time spent in analyzed innermost loops (12.18%)
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 (96.39%)
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 (0.00%)
[ 3 / 3 ] Cumulative Outermost/In between loops coverage (7.90%) lower than cumulative innermost loop coverage (12.18%)
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% (0.00%) is spend in Libm/SVML (special functions)
| Loop ID | Analysis | Penalty Score |
|---|---|---|
| ►Loop 604 - libdmumps.so | Execution Time: 7 % - Vectorization Ratio: 5.13 % - Vector Length Use: 12.20 % | |
| ►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 (Outermost) - 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 (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ○Loop 605 - libdmumps.so | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 2086 - libdmumps.so | Execution Time: 2 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 |
| ►Data Access Issues | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ►Vectorization Roadblocks | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ○Loop 2286 - libdmumps.so | Execution Time: 1 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 2282 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►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 | 2 | |
| ○ | [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 |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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] 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 |
| ○Loop 2294 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 2194 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 22.51 % - Vector Length Use: 14.07 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 | 5 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 3 issues (= calls) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Data Access Issues | 3 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT) - Simplify data access and try to get stride 1 access. There are 1 issues (= instructions) costing 1 point each. | 1 |
| ○ | [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 | 1005 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 3 issues (= calls) costing 1 point each. | 3 |
| ○ | [SA] Too many paths (at least 1000 paths) - Simplify control structure. There are at least 1000 issues ( = paths) costing 1 point. | 1000 |
| ○ | [SA] Non innermost loop (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Inefficient Vectorization | 1 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT) - Simplify data access and try to get stride 1 access. There are 1 issues (= instructions) costing 1 point each. | 1 |
| ►Loop 609 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 33.33 % - Vector Length Use: 16.67 % | |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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 |
| ►Loop 2502 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 12.70 % - Vector Length Use: 13.29 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 | 2 | |
| ○ | [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 |
| ►Data Access Issues | 8 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 2 issues ( = indirect data accesses) costing 4 point each. | 8 |
| ►Vectorization Roadblocks | 10 | |
| ○ | [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 indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 2 issues ( = indirect data accesses) costing 4 point each. | 8 |
| ○Loop 2482 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % |
[ 4 / 4 ] Application profile is long enough (329.60 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with source/debug info
-g option gives access to debugging informations, such are source locations.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with compilation options informations and -fno-omit-frame-pointer is present
-fno-omit-frame-pointer improves the accuracy of callchains found during the application profiling.
[ 3 / 3 ] Host configuration allows retrieval of all necessary metrics.
[ 0 / 3 ] Compilation of some functions is not optimized for the target processor
Application run on the GRANITE_RAPIDS micro-architecture while the code was specialized for sapphirerapids.
Architecture specific options are needed to produce efficient code for a specific processor ( -march=(target) ).
[ 3 / 3 ] Optimization level option is correctly used
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 0.00 % 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.
[ 0 / 4 ] Too little time of the experiment time spent in analyzed loops (19.75%)
If the time spent in analyzed loops is less than 30%, standard loop optimizations will have a limited impact on application performances.
[ 2 / 4 ] A significant amount of threads are idle (39.68%)
On average, more than 10% of observed threads are idle. Such threads are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 2 / 4 ] CPU activity is below 90% (61.68%)
CPU cores are idle more than 10% of time. Threads supposed to run on these cores are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 4 / 4 ] Loop profile is not flat
At least one loop coverage is greater than 4% (7.12%), representing an hotspot for the application
[ 0 / 4 ] Too little time of the experiment time spent in analyzed innermost loops (13.14%)
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 (97.73%)
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 (1.56%)
[ 3 / 3 ] Cumulative Outermost/In between loops coverage (6.61%) lower than cumulative innermost loop coverage (13.14%)
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% (0.00%) is spend in Libm/SVML (special functions)
| Loop ID | Analysis | Penalty Score |
|---|---|---|
| ○Loop 605 - libdmumps.so | Execution Time: 7 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 604 - libdmumps.so | Execution Time: 6 % - Vectorization Ratio: 5.13 % - Vector Length Use: 12.20 % | |
| ►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 (Outermost) - 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 (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 2086 - libdmumps.so | Execution Time: 2 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 |
| ►Data Access Issues | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ►Vectorization Roadblocks | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ○Loop 2286 - libdmumps.so | Execution Time: 1 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 2282 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►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 | 2 | |
| ○ | [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 |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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] 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 |
| ►Loop 949 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Data Access Issues | 4 | |
| ○ | [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 |
| ►Vectorization Roadblocks | 4 | |
| ○ | [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 |
| ○Loop 2287 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 2194 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 22.51 % - Vector Length Use: 14.07 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 | 5 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 3 issues (= calls) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Data Access Issues | 3 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT) - Simplify data access and try to get stride 1 access. There are 1 issues (= instructions) costing 1 point each. | 1 |
| ○ | [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 | 1005 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 3 issues (= calls) costing 1 point each. | 3 |
| ○ | [SA] Too many paths (at least 1000 paths) - Simplify control structure. There are at least 1000 issues ( = paths) costing 1 point. | 1000 |
| ○ | [SA] Non innermost loop (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Inefficient Vectorization | 1 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT) - Simplify data access and try to get stride 1 access. There are 1 issues (= instructions) costing 1 point each. | 1 |
| ►Loop 935 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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 |
| ○Loop 2294 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % |
[ 4 / 4 ] Application profile is long enough (265.11 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with source/debug info
-g option gives access to debugging informations, such are source locations.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with compilation options informations and -fno-omit-frame-pointer is present
-fno-omit-frame-pointer improves the accuracy of callchains found during the application profiling.
[ 3 / 3 ] Host configuration allows retrieval of all necessary metrics.
[ 0 / 3 ] Compilation of some functions is not optimized for the target processor
Application run on the GRANITE_RAPIDS micro-architecture while the code was specialized for sapphirerapids.
Architecture specific options are needed to produce efficient code for a specific processor ( -march=(target) ).
[ 3 / 3 ] Optimization level option is correctly used
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 0.00 % 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.
[ 0 / 4 ] Too little time of the experiment time spent in analyzed loops (17.22%)
If the time spent in analyzed loops is less than 30%, standard loop optimizations will have a limited impact on application performances.
[ 1 / 4 ] A significant amount of threads are idle (62.25%)
On average, more than 10% of observed threads are idle. Such threads are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 1 / 4 ] CPU activity is below 90% (39.36%)
CPU cores are idle more than 10% of time. Threads supposed to run on these cores are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 4 / 4 ] Loop profile is not flat
At least one loop coverage is greater than 4% (6.25%), representing an hotspot for the application
[ 0 / 4 ] Too little time of the experiment time spent in analyzed innermost loops (12.57%)
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 (98.55%)
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 (2.28%)
[ 3 / 3 ] Cumulative Outermost/In between loops coverage (4.65%) lower than cumulative innermost loop coverage (12.57%)
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% (0.00%) is spend in Libm/SVML (special functions)
| Loop ID | Analysis | Penalty Score |
|---|---|---|
| ○Loop 605 - libdmumps.so | Execution Time: 6 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 604 - libdmumps.so | Execution Time: 4 % - Vectorization Ratio: 5.13 % - Vector Length Use: 12.20 % | |
| ►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 (Outermost) - 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 (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 2086 - libdmumps.so | Execution Time: 2 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 |
| ►Data Access Issues | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ►Vectorization Roadblocks | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ○Loop 2286 - libdmumps.so | Execution Time: 1 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 949 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Data Access Issues | 4 | |
| ○ | [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 |
| ►Vectorization Roadblocks | 4 | |
| ○ | [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 |
| ►Loop 2282 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►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 | 2 | |
| ○ | [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 |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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] 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 |
| ○Loop 2287 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 2194 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 22.51 % - Vector Length Use: 14.07 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 | 5 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 3 issues (= calls) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Data Access Issues | 3 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT) - Simplify data access and try to get stride 1 access. There are 1 issues (= instructions) costing 1 point each. | 1 |
| ○ | [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 | 1005 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 3 issues (= calls) costing 1 point each. | 3 |
| ○ | [SA] Too many paths (at least 1000 paths) - Simplify control structure. There are at least 1000 issues ( = paths) costing 1 point. | 1000 |
| ○ | [SA] Non innermost loop (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Inefficient Vectorization | 1 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT) - Simplify data access and try to get stride 1 access. There are 1 issues (= instructions) costing 1 point each. | 1 |
| ►Loop 935 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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 |
| ○Loop 2294 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % |
[ 4 / 4 ] Application profile is long enough (256.14 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with source/debug info
-g option gives access to debugging informations, such are source locations.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with compilation options informations and -fno-omit-frame-pointer is present
-fno-omit-frame-pointer improves the accuracy of callchains found during the application profiling.
[ 3 / 3 ] Host configuration allows retrieval of all necessary metrics.
[ 0 / 3 ] Compilation of some functions is not optimized for the target processor
Application run on the GRANITE_RAPIDS micro-architecture while the code was specialized for sapphirerapids.
Architecture specific options are needed to produce efficient code for a specific processor ( -march=(target) ).
[ 3 / 3 ] Optimization level option is correctly used
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 0.00 % 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.
[ 0 / 4 ] Too little time of the experiment time spent in analyzed loops (15.53%)
If the time spent in analyzed loops is less than 30%, standard loop optimizations will have a limited impact on application performances.
[ 0 / 4 ] A significant amount of threads are idle (78.93%)
On average, more than 10% of observed threads are idle. Such threads are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 0 / 4 ] CPU activity is below 90% (22.17%)
CPU cores are idle more than 10% of time. Threads supposed to run on these cores are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 4 / 4 ] Loop profile is not flat
At least one loop coverage is greater than 4% (5.37%), representing an hotspot for the application
[ 0 / 4 ] Too little time of the experiment time spent in analyzed innermost loops (11.98%)
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 (95.66%)
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 (2.64%)
[ 3 / 3 ] Cumulative Outermost/In between loops coverage (3.56%) lower than cumulative innermost loop coverage (11.98%)
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% (0.00%) is spend in Libm/SVML (special functions)
| Loop ID | Analysis | Penalty Score |
|---|---|---|
| ○Loop 605 - libdmumps.so | Execution Time: 5 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 604 - libdmumps.so | Execution Time: 3 % - Vectorization Ratio: 5.13 % - Vector Length Use: 12.20 % | |
| ►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 (Outermost) - 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 (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 2086 - libdmumps.so | Execution Time: 2 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 |
| ►Data Access Issues | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ►Vectorization Roadblocks | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ○Loop 2286 - libdmumps.so | Execution Time: 2 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 2282 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►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 | 2 | |
| ○ | [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 |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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] 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 |
| ►Loop 949 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Data Access Issues | 4 | |
| ○ | [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 |
| ►Vectorization Roadblocks | 4 | |
| ○ | [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 |
| ○Loop 2287 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 935 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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 |
| ►Loop 2194 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 22.51 % - Vector Length Use: 14.07 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 | 5 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 3 issues (= calls) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Data Access Issues | 3 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT) - Simplify data access and try to get stride 1 access. There are 1 issues (= instructions) costing 1 point each. | 1 |
| ○ | [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 | 1005 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 3 issues (= calls) costing 1 point each. | 3 |
| ○ | [SA] Too many paths (at least 1000 paths) - Simplify control structure. There are at least 1000 issues ( = paths) costing 1 point. | 1000 |
| ○ | [SA] Non innermost loop (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Inefficient Vectorization | 1 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT) - Simplify data access and try to get stride 1 access. There are 1 issues (= instructions) costing 1 point each. | 1 |
| ○Loop 2294 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % |
[ 4 / 4 ] Application profile is long enough (293.28 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with source/debug info
-g option gives access to debugging informations, such are source locations.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with compilation options informations and -fno-omit-frame-pointer is present
-fno-omit-frame-pointer improves the accuracy of callchains found during the application profiling.
[ 3 / 3 ] Host configuration allows retrieval of all necessary metrics.
[ 0 / 3 ] Compilation of some functions is not optimized for the target processor
Application run on the GRANITE_RAPIDS micro-architecture while the code was specialized for sapphirerapids.
Architecture specific options are needed to produce efficient code for a specific processor ( -march=(target) ).
[ 3 / 3 ] Optimization level option is correctly used
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 0.00 % 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.
[ 0 / 4 ] Too little time of the experiment time spent in analyzed loops (13.00%)
If the time spent in analyzed loops is less than 30%, standard loop optimizations will have a limited impact on application performances.
[ 0 / 4 ] A significant amount of threads are idle (88.82%)
On average, more than 10% of observed threads are idle. Such threads are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 0 / 4 ] CPU activity is below 90% (11.73%)
CPU cores are idle more than 10% of time. Threads supposed to run on these cores are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 0 / 4 ] Loop profile is flat
No hotspot found in the application (greatest loop coverage is 3.97%), and the twenty hottest loops cumulated coverage is lower than 20% of the application profiled time (12.92%)
[ 0 / 4 ] Too little time of the experiment time spent in analyzed innermost loops (11.18%)
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 (99.62%)
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 (3.03%)
[ 3 / 3 ] Cumulative Outermost/In between loops coverage (1.82%) lower than cumulative innermost loop coverage (11.18%)
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% (0.00%) is spend in Libm/SVML (special functions)
| Loop ID | Analysis | Penalty Score |
|---|---|---|
| ○Loop 605 - libdmumps.so | Execution Time: 3 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 2086 - libdmumps.so | Execution Time: 3 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 |
| ►Data Access Issues | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ►Vectorization Roadblocks | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ○Loop 2286 - libdmumps.so | Execution Time: 2 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 604 - libdmumps.so | Execution Time: 1 % - Vectorization Ratio: 5.13 % - Vector Length Use: 12.20 % | |
| ►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 (Outermost) - 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 (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 2282 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►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 | 2 | |
| ○ | [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 |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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] 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 |
| ►Loop 949 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Data Access Issues | 4 | |
| ○ | [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 |
| ►Vectorization Roadblocks | 4 | |
| ○ | [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 |
| ○Loop 2287 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 935 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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 |
| ►Loop 2194 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 22.51 % - Vector Length Use: 14.07 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 | 5 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 3 issues (= calls) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Data Access Issues | 3 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT) - Simplify data access and try to get stride 1 access. There are 1 issues (= instructions) costing 1 point each. | 1 |
| ○ | [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 | 1005 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 3 issues (= calls) costing 1 point each. | 3 |
| ○ | [SA] Too many paths (at least 1000 paths) - Simplify control structure. There are at least 1000 issues ( = paths) costing 1 point. | 1000 |
| ○ | [SA] Non innermost loop (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Inefficient Vectorization | 1 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT) - Simplify data access and try to get stride 1 access. There are 1 issues (= instructions) costing 1 point each. | 1 |
| ○Loop 2294 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % |
[ 4 / 4 ] Application profile is long enough (310.53 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with source/debug info
-g option gives access to debugging informations, such are source locations.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with compilation options informations and -fno-omit-frame-pointer is present
-fno-omit-frame-pointer improves the accuracy of callchains found during the application profiling.
[ 3 / 3 ] Host configuration allows retrieval of all necessary metrics.
[ 0 / 3 ] Compilation of some functions is not optimized for the target processor
Application run on the GRANITE_RAPIDS micro-architecture while the code was specialized for sapphirerapids.
Architecture specific options are needed to produce efficient code for a specific processor ( -march=(target) ).
[ 3 / 3 ] Optimization level option is correctly used
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 0.00 % 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.
[ 0 / 4 ] Too little time of the experiment time spent in analyzed loops (11.40%)
If the time spent in analyzed loops is less than 30%, standard loop optimizations will have a limited impact on application performances.
[ 0 / 4 ] A significant amount of threads are idle (91.57%)
On average, more than 10% of observed threads are idle. Such threads are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 0 / 4 ] CPU activity is below 90% (8.84%)
CPU cores are idle more than 10% of time. Threads supposed to run on these cores are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 0 / 4 ] Loop profile is flat
No hotspot found in the application (greatest loop coverage is 3.23%), and the twenty hottest loops cumulated coverage is lower than 20% of the application profiled time (11.34%)
[ 0 / 4 ] Too little time of the experiment time spent in analyzed innermost loops (10.19%)
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 (99.80%)
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 (4.12%)
[ 3 / 3 ] Cumulative Outermost/In between loops coverage (1.22%) lower than cumulative innermost loop coverage (10.19%)
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% (0.00%) is spend in Libm/SVML (special functions)
| Loop ID | Analysis | Penalty Score |
|---|---|---|
| ►Loop 2086 - libdmumps.so | Execution Time: 3 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 |
| ►Data Access Issues | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ►Vectorization Roadblocks | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ○Loop 605 - libdmumps.so | Execution Time: 2 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ○Loop 2286 - libdmumps.so | Execution Time: 2 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 604 - libdmumps.so | Execution Time: 1 % - Vectorization Ratio: 5.13 % - Vector Length Use: 12.20 % | |
| ►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 (Outermost) - 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 (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 2282 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►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 | 2 | |
| ○ | [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 |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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] 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 |
| ►Loop 949 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Data Access Issues | 4 | |
| ○ | [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 |
| ►Vectorization Roadblocks | 4 | |
| ○ | [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 |
| ○Loop 603 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 935 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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 |
| ○Loop 2287 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 2194 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 22.51 % - Vector Length Use: 14.07 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 | 5 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 3 issues (= calls) costing 1 point each. | 3 |
| ○ | [SA] Non innermost loop (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Data Access Issues | 3 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT) - Simplify data access and try to get stride 1 access. There are 1 issues (= instructions) costing 1 point each. | 1 |
| ○ | [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 | 1005 | |
| ○ | [SA] Presence of calls - Inline either by compiler or by hand and use SVML for libm calls. There are 3 issues (= calls) costing 1 point each. | 3 |
| ○ | [SA] Too many paths (at least 1000 paths) - Simplify control structure. There are at least 1000 issues ( = paths) costing 1 point. | 1000 |
| ○ | [SA] Non innermost loop (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Inefficient Vectorization | 1 | |
| ○ | [SA] Presence of special instructions executing on a single port (INSERT/EXTRACT) - Simplify data access and try to get stride 1 access. There are 1 issues (= instructions) costing 1 point each. | 1 |
[ 4 / 4 ] Application profile is long enough (325.09 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with source/debug info
-g option gives access to debugging informations, such are source locations.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with compilation options informations and -fno-omit-frame-pointer is present
-fno-omit-frame-pointer improves the accuracy of callchains found during the application profiling.
[ 3 / 3 ] Host configuration allows retrieval of all necessary metrics.
[ 0 / 3 ] Compilation of some functions is not optimized for the target processor
Application run on the GRANITE_RAPIDS micro-architecture while the code was specialized for sapphirerapids.
Architecture specific options are needed to produce efficient code for a specific processor ( -march=(target) ).
[ 3 / 3 ] Optimization level option is correctly used
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 0.00 % 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.
[ 0 / 4 ] Too little time of the experiment time spent in analyzed loops (7.86%)
If the time spent in analyzed loops is less than 30%, standard loop optimizations will have a limited impact on application performances.
[ 0 / 4 ] A significant amount of threads are idle (92.22%)
On average, more than 10% of observed threads are idle. Such threads are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 0 / 4 ] CPU activity is below 90% (8.14%)
CPU cores are idle more than 10% of time. Threads supposed to run on these cores are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 0 / 4 ] Loop profile is flat
No hotspot found in the application (greatest loop coverage is 1.94%), and the twenty hottest loops cumulated coverage is lower than 20% of the application profiled time (7.82%)
[ 0 / 4 ] Too little time of the experiment time spent in analyzed innermost loops (7.01%)
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 (99.88%)
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 (1.50%)
[ 3 / 3 ] Cumulative Outermost/In between loops coverage (0.85%) lower than cumulative innermost loop coverage (7.01%)
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% (0.00%) is spend in Libm/SVML (special functions)
| Loop ID | Analysis | Penalty Score |
|---|---|---|
| ○Loop 605 - libdmumps.so | Execution Time: 1 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 2086 - libdmumps.so | Execution Time: 1 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 |
| ►Data Access Issues | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ►Vectorization Roadblocks | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ○Loop 603 - libdmumps.so | Execution Time: 1 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ○Loop 2286 - libdmumps.so | Execution Time: 1 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 604 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 5.13 % - Vector Length Use: 12.20 % | |
| ►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 (Outermost) - 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 (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 2282 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►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 | 2 | |
| ○ | [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 |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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] 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 |
| ►Loop 949 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Data Access Issues | 4 | |
| ○ | [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 |
| ►Vectorization Roadblocks | 4 | |
| ○ | [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 |
| ►Loop 935 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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 |
| ○Loop 2287 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 609 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 33.33 % - Vector Length Use: 16.67 % | |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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 |
[ 4 / 4 ] Application profile is long enough (395.86 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with source/debug info
-g option gives access to debugging informations, such are source locations.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with compilation options informations and -fno-omit-frame-pointer is present
-fno-omit-frame-pointer improves the accuracy of callchains found during the application profiling.
[ 3 / 3 ] Host configuration allows retrieval of all necessary metrics.
[ 0 / 3 ] Compilation of some functions is not optimized for the target processor
Application run on the GRANITE_RAPIDS micro-architecture while the code was specialized for sapphirerapids.
Architecture specific options are needed to produce efficient code for a specific processor ( -march=(target) ).
[ 3 / 3 ] Optimization level option is correctly used
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 0.00 % 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.
[ 0 / 4 ] Too little time of the experiment time spent in analyzed loops (6.71%)
If the time spent in analyzed loops is less than 30%, standard loop optimizations will have a limited impact on application performances.
[ 0 / 4 ] A significant amount of threads are idle (84.43%)
On average, more than 10% of observed threads are idle. Such threads are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 0 / 4 ] CPU activity is below 90% (16.23%)
CPU cores are idle more than 10% of time. Threads supposed to run on these cores are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 0 / 4 ] Loop profile is flat
No hotspot found in the application (greatest loop coverage is 3.39%), and the twenty hottest loops cumulated coverage is lower than 20% of the application profiled time (6.71%)
[ 0 / 4 ] Too little time of the experiment time spent in analyzed innermost loops (6.08%)
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 (99.90%)
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
[ 0 / 3 ] Too many functions do not use all threads
Functions running on a reduced number of threads (typically sequential code) cover at least 10% of application walltime (11.68%). Check both "Max Inclusive Time Over Threads" and "Nb Threads" in Functions or Loops tabs and consider parallelizing sequential regions or improving parallelization of regions running on a reduced number of threads
[ 3 / 3 ] Cumulative Outermost/In between loops coverage (0.63%) lower than cumulative innermost loop coverage (6.08%)
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% (0.00%) is spend in Libm/SVML (special functions)
| Loop ID | Analysis | Penalty Score |
|---|---|---|
| ○Loop 603 - libdmumps.so | Execution Time: 3 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ○Loop 605 - libdmumps.so | Execution Time: 1 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 604 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 5.13 % - Vector Length Use: 12.20 % | |
| ►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 (Outermost) - 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 (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 2086 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 |
| ►Data Access Issues | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ►Vectorization Roadblocks | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ○Loop 2286 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 2282 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►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 | 2 | |
| ○ | [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 |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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] 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 |
| ○Loop 634 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ○Loop 608 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ○Loop 635 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 609 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 33.33 % - Vector Length Use: 16.67 % | |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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 |
[ 4 / 4 ] Application profile is long enough (466.05 s)
To have good quality measurements, it is advised that the application profiling time is greater than 10 seconds.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with source/debug info
-g option gives access to debugging informations, such are source locations.
[ 3 / 3 ] Most of time spent in analyzed modules comes from functions with compilation options informations and -fno-omit-frame-pointer is present
-fno-omit-frame-pointer improves the accuracy of callchains found during the application profiling.
[ 3 / 3 ] Host configuration allows retrieval of all necessary metrics.
[ 0 / 3 ] Compilation of some functions is not optimized for the target processor
Application run on the GRANITE_RAPIDS micro-architecture while the code was specialized for sapphirerapids.
Architecture specific options are needed to produce efficient code for a specific processor ( -march=(target) ).
[ 3 / 3 ] Optimization level option is correctly used
[ 2 / 2 ] Application is correctly profiled ("Others" category represents 0.00 % 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.
[ 0 / 4 ] Too little time of the experiment time spent in analyzed loops (5.69%)
If the time spent in analyzed loops is less than 30%, standard loop optimizations will have a limited impact on application performances.
[ 0 / 4 ] A significant amount of threads are idle (79.31%)
On average, more than 10% of observed threads are idle. Such threads are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 0 / 4 ] CPU activity is below 90% (21.40%)
CPU cores are idle more than 10% of time. Threads supposed to run on these cores are probably IO/sync waiting. Some hints: use faster filesystems to read/write data, improve parallel load balancing and/or scheduling.
[ 0 / 4 ] Loop profile is flat
No hotspot found in the application (greatest loop coverage is 3.18%), and the twenty hottest loops cumulated coverage is lower than 20% of the application profiled time (5.69%)
[ 0 / 4 ] Too little time of the experiment time spent in analyzed innermost loops (5.16%)
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 (99.94%)
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
[ 0 / 3 ] Too many functions do not use all threads
Functions running on a reduced number of threads (typically sequential code) cover at least 10% of application walltime (15.36%). Check both "Max Inclusive Time Over Threads" and "Nb Threads" in Functions or Loops tabs and consider parallelizing sequential regions or improving parallelization of regions running on a reduced number of threads
[ 3 / 3 ] Cumulative Outermost/In between loops coverage (0.53%) lower than cumulative innermost loop coverage (5.16%)
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% (0.00%) is spend in Libm/SVML (special functions)
| Loop ID | Analysis | Penalty Score |
|---|---|---|
| ○Loop 603 - libdmumps.so | Execution Time: 3 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ○Loop 605 - libdmumps.so | Execution Time: 1 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 604 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 5.13 % - Vector Length Use: 12.20 % | |
| ►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 (Outermost) - 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 (Outermost) - Collapse loop with innermost ones. This issue costs 2 points. | 2 |
| ►Loop 2086 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop Computation Issues | 6 | |
| ○ | [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 |
| ○ | [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 |
| ►Data Access Issues | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ►Vectorization Roadblocks | 4 | |
| ○ | [SA] Presence of indirect accesses - Use array restructuring or gather instructions to lower the cost. There are 1 issues ( = indirect data accesses) costing 4 point each. | 4 |
| ○Loop 2286 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 2282 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►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 | 2 | |
| ○ | [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 |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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] 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 |
| ○Loop 634 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ○Loop 608 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ○Loop 635 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 0.00 % - Vector Length Use: 12.50 % | |
| ►Loop 609 - libdmumps.so | Execution Time: 0 % - Vectorization Ratio: 33.33 % - Vector Length Use: 16.67 % | |
| ►Data Access Issues | 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 |
| ►Vectorization Roadblocks | 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 |