CUDA ISP for NVIDIA Jetson/Performance: Difference between revisions

CUDA ISP for NVIDIA Jetson/Performance (view source)

Revision as of 13:40, 24 March 2023

1,108 bytes added , 24 March 2023

no edit summary

Fguindon

183

edits

@@ Line 4: / Line 4: @@
 {{DISPLAYTITLE:CUDA ISP for NVIDIA Jetson: Performance|noerror}}
+= Library API performance =
+To measure the CUDA ISP API performance, we built a simple example that iterates over the apply methods and records performance metrics for each iteration. We recorded the duration of each apply method, the CPU and GPU usage during the application of the code, and the CPU RAM and GPU RAM usage. We recorded the performance statistics over 3 buffer sizes. We recorded the performance on a Jetson Nano, Jetson Xavier NX, Jetson Xavier AGX, and Jetson Orin.
+* A minimum 2x2 case, to test the maximum speeds that the apply methods could achieve
+* A medium 1920x1080 case, to illustrate the changes in performance as the buffer size increases
+* A maximum 3840x2160 case, to test performance on large buffers
+== Procesing time ==
+<center>
+{| class="wikitable"
+|-
+! Procesing time (In microseconds, averaged over 100 iterations) !! Jetson Nano !! Jetson Xavier NX !! Jetson Xavier AGX !! Jetson Orin
+|-
+| cudashift || 396
+|-
+| cudadebayer || 0.002522
+|-
+| cudawhitebalancer ||
+|-
+| cudacolorspaceconverter ||
+|-
+|}
+</center>
+== CPU and CPU RAM usage ==
+== GPU and GPU RAM usage ==
 = GStreamer elements performance =
@@ Line 9: / Line 37: @@
 For testing purposes, take into account the following points:
-* Maximun performance mode enabled: all cores and Jetson clocks enabled.
+* Maximum performance mode enabled: all cores and Jetson clocks enabled.
 * Jetpack 4.6
 * FPS is equal to 1/processing time