General

The Raspberry Pi Compute Module 4 is a well priced platform regarding its performance.

CPU	Cortex-A72 @ 1500MHz
Cores	4
RAM	4 GB
USB2.0 Interfaces	1
Ethernet	10/100/1000 MBit
PCIe	1 x 1 Lane Gen 2.0

Note: The above table describes the specification of the 4GB version Raspberry Pi Compute Module 4.; The Raspberry Pi Compute Module 4 IO Board is used as carrier board for the following test.

Test Setup

Test setup

Note: Since the Raspberry Pi Compute Module 4 only supports single-lane PCIe 2.0, the edge of the slot needs to be cut off in order to insert the BVS CA-BN adapter board (like shown in the test setup above). Besides, the max. throughput of the device is limited to < 400MB/s due to the 1x PCIe lane width. For Raspberry Pi OS (64-bit) on the Raspberry Pi Compute Module 4, please make sure that the Kernel version is >=5.10.11.

Additional Settings

In order for the system's X Server not to auto-detect the BVS CA-BN as a GPU, the following configuration has to be adapted:

Create a directory /etc/X11/xorg.conf.d.
Create a file named 10-force_no_auto.conf under the directory /etc/X11/xorg.conf.d.

Added the following content to the file:

Section "ServerFlags"
        Option "AutoAddGPU" "false"
EndSection

Benchmarks

The following tests have been performed using different de-Bayering scenarios to achieve the max. FPS while maintaining 0 lost images. The CPU load during the acquisition is also documented below.

Scenarios that have been tested are listed as follows:

When de-Bayering is carried out on the camera: The camera delivers RGB8 image data to the host system. This setting results in a lower CPU load but a lower frame rate.
When de-Bayering is carried out on the host system: The camera delivers Bayer8 image data to the host system. The Bayer8 image data then get de-Bayered to RGB8 format on the host system. This setting results in a higher frame rate but a higher CPU load as well.
When no de-Bayering is performed: The camera delivers Bayer8 image data to the host system. No de-Bayering is performed. This settings results in a lower CPU load and a higher frame rate. The behavior is identical to monochrome cameras.

Camera	Resolution	Pixel Format	Frame Rate [Frames/s]	Bandwidth [MB/s]	CPU Load With Image Display	CPU Load Without Image Display
CA-BN2-0051ZC	2464 x 2056	RGB8 (on camera) → RGB8 (on host)	24.70	375.47	~8%	~4%


Camera	Resolution	Pixel Format	Frame Rate [Frames/s]	Bandwidth [MB/s]	CPU Load With Image Display	CPU Load Without Image Display
CA-BN2-0051ZC	2464 x 2056	BayerRG8 (on camera) → RGB8 (on host)	56.90	288.29	~72%	~67%


Camera	Resolution	Pixel Format	Frame Rate [Frames/s]	Bandwidth [MB/s]	CPU Load With Image Display	CPU Load Without Image Display
CA-BN2-0051ZC	2464 x 2056	BayerRG8 (on camera) → BayerRG8/Raw (on host)	74.20	375.94	~14%	~8%

Note: Since performing De-Bayering on the host (e.g. scenario 2) requires a lot more CPU processing power, the gross transmission bandwidth (i.e. DeviceLinkThroughput) has been limited to 330MB/s to avoid frame loss during acquisition.