ArithmeticUtilization (Time Consumptions and Percentages of Cube and Vector Instructions)

ArithmeticUtilization.csv stores the cycle ratio data of Cube and Vector instructions. You are advised to optimize the operator logic to reduce redundant computing instructions. For details, see the field description in the following table.

Figure 1 ArithmeticUtilization.csv file

See the following table for more details.

**Table 1** Field description
Field	Description
block_id	Number of running task blocks, which corresponds to the number of cores configured during task running.
sub_block_id	ID of the Vector/Cube core on the block.
aic_time(us)	Execution time (μs) on each AI Cube Core compute unit after the task is allocated to each AI Cube Core compute unit.
aic_total_cycles	Total number of cycles executed on each AI Cube Core compute unit after the task is allocated to each AI Cube Core compute unit.
aiv_time(us)	Execution time of each AI Vector Core compute unit after the task is allocated to the unit, in μs.
aiv_total_cycles	Total number of cycles executed on each AI Vector Core compute unit after the task is allocated to the unit.
aic_cube_ratio	Ratio of cycles taken to execute Cube instructions to the total cycles.
aic_cube_fp16_ratio	Ratio of cycles taken to execute Cube fp16 instructions to the total cycles.
aic_cube_int8_ratio	Ratio of cycles taken to execute Cube int8 instructions to the total cycles.
aic_cube_fops	Floating-point operations (fops in this field) of the Cube type, indicating the computation amount. This field can be used to measure the complexity of an algorithm or model.
aic_cube_total_instr_number	Total number of Cube instructions, including the fp and int types.
aic_cube_fp_instr_number	Total number of Cube fp instructions.
aic_cube_int_instr_number	Total number of Cube int instructions.
aiv_vec_ratio	Ratio of cycles taken to execute Vector instructions to the total cycles.
aiv_vec_fp32_ratio	Ratio of cycles taken to execute Vector fp32 instructions to the total cycles.
aiv_vec_fp16_ratio	Ratio of cycles taken to execute Vector fp16 instructions to the total cycles.
aiv_vec_int32_ratio	Ratio of cycles taken to execute Vector int32 instructions to the total cycles.
aiv_vec_int16_ratio	Ratio of cycles taken to execute Vector int16 instructions to the total cycles.
aiv_vec_misc_ratio	Ratio of cycles taken to execute Vector misc instructions to the total cycles.
aiv_vec_fops	Floating-point operations (fops in this field) of the Vector type, indicating the computation amount. This field can be used to measure the complexity of an algorithm or model.

Figure 2 ArithmeticUtilization.csv file

See the following table for more details.

**Table 2** Field description
Field	Description
aic_time(us)	Execution time of each AI Core compute unit after the task is allocated to the unit, in μs.
aic_total_cycles	Total number of cycles executed on each AI Core compute unit after the task is allocated to the unit.
aic_cube_ratio	Ratio of cycles taken to execute Cube instructions to the total cycles.
aic_cube_fp16_ratio	Ratio of cycles taken to execute Cube fp16 instructions to the total cycles.
aic_cube_int8_ratio	Ratio of cycles taken to execute Cube int8 instructions to the total cycles.
aic_cube_fops	Floating-point operations (fops in this field) of the Cube type, indicating the computation amount. This field can be used to measure the complexity of an algorithm or model.
aic_cube_total_instr_number	Total number of Cube instructions, including the fp and int types.
aic_vec_ratio	Ratio of cycles taken to execute Vector instructions to the total cycles.
aic_vec_fp32_ratio	Ratio of cycles taken to execute Vector fp32 instructions to the total cycles.
aic_vec_fp16_ratio	Ratio of cycles taken to execute Vector fp16 instructions to the total cycles.
aic_vec_int32_ratio	Ratio of cycles taken to execute Vector int32 instructions to the total cycles.
aic_vec_ int16_ratio	Ratio of cycles taken to execute Vector int16 instructions to the total cycles.
aic_vec_misc_ratio	Ratio of cycles taken to execute Vector misc instructions to the total cycles.
aic_vec_fops	Floating-point operations (fops in this field) of the Vector type, indicating the computation amount. This field can be used to measure the complexity of an algorithm or model.

Parent topic: msprof op