Intel^® VTune^™ Profiler 2022.2.0

EU Array Stalled/Idle: 25.6% of Elapsed time with GPU busy

The percentage of time when the EUs were stalled or idle is high, which has a negative impact on compute-bound applications.

GPU L3 Bandwidth Bound: 97.3% of peak value

L3 bandwidth was high when EUs were stalled or idle. Consider improving cache reuse.

Hottest GPU Computing Tasks Bound by GPU L3 Bandwidth:

Computing Task	Total Time
dppyPy_dppy_py_devfn__5F__5F_main_5F__5F__2E_run_5F_knn_5F_kernel_24_1_2E_array_28_float64_2C__20_2d_2C__20_C_29__2E_array_28_int64_2C__20_1d_2C__20_C_29__2E_array_28_float64_2C__20_2d_2C__20_C_29__2E_int64_2E_int64_2E_int64_2E_array_28_float64_2C__20_1d_2C__20_C_29__2E_array_28_float64_2C__20_2d_2C__20_C_29__2E_int64	195.403s

Sampler Busy: 0.0% of peak value
- Hottest GPU Computing Tasks with High Sampler Usage:
  Computing Task Total Time

FPU Utilization: 26.8% of Elapsed time with GPU busy
- Hottest GPU Computing Tasks with High FPU Utilization:
  Computing Task Total Time