ximea
XIMEA PCIe cameras connect directly to the host over PCI Express – up to 64 Gbit/s over PCIe Gen 3. The cameras support DMA transfer to memory, and data transfer to the GPU is possible on Linux, which keeps latency low. A major benefit is the ability to easily combine these cameras with SoC (system on a chip) modules with a GPU, like the ones from NVIDIA.
The NVIDIA Orin modules opened up new performance levels for using XIMEA high-speed, high-resolution PCIe cameras in embedded vision applications. Together with the XIMEA XEC-NX-3P-X2G3 carrier board, up to three cameras can be attached to a single Orin NX module.
One, two and three XIMEA cameras were tested together with the Jetson Orin NX module by XIMEA's technology partner MRTech. Below is a description of the test setup, which streams images from the cameras to a computer, together with the results.
Highlights: even this compact module streamed Full HD video at 390 fps with a glass-to-glass latency of only 28 ms, and ran three cameras at 120 fps each with an average latency of 43 ms – all within a power budget of around 40 W including the cameras. This makes the Orin NX a good fit for small multi-camera systems in drones, robots, vehicles and medical devices.
Picture 1. Setup with three XIMEA cameras, Orin NX and the XIMEA XEC-NX-3P-X2G3 carrier board
The NVIDIA Jetson Orin NX is a compact module of the Jetson Orin family in the small SO-DIMM form factor. It has an NVIDIA Ampere architecture GPU with Tensor Cores, a multi-core Arm CPU, a hardware video encoder and PCIe Gen 4 lanes for attaching cameras. With JetPack 6.2 the module supports the MAXN Super power mode, which raises the GPU and CPU clocks and the AI performance compared to the original power modes. The module itself is smaller than a credit card, and with the XIMEA carrier board it becomes a small multi-camera platform with up to three cameras.
Table 1. Jetson Orin NX module variants (NVIDIA specification, with Super power modes)
| Parameter | Orin NX 8GB | Orin NX 16GB |
| AI performance | 117 TOPS | 157 TOPS |
| GPU | 1024-core Ampere, 32 Tensor Cores | 1024-core Ampere, 32 Tensor Cores |
| CPU | 6-core Arm Cortex-A78AE | 8-core Arm Cortex-A78AE |
| Memory | 8 GB 128-bit LPDDR5, 102.4 GB/s | 16 GB 128-bit LPDDR5, 102.4 GB/s |
| Video encode | 1× 4K60 H.265 | 1× 4K60 H.265 |
| PCIe | 1× x4 + 3× x1 (PCIe Gen 4) | 1× x4 + 3× x1 (PCIe Gen 4) |
| Module power | 10 W – 40 W (MAXN Super) | 10 W – 40 W (MAXN Super) |
In these tests the module was mounted on the XIMEA XEC-NX-3P-X2G3 carrier board. The board can be configured either with three PCIe ports for X2G2 cameras or with one port for an X4G3 camera, and the cameras are connected over FireFly flat cables. It was used in its 1× X4G3 configuration for the single MX245CG camera and in its 3× X2G2 configuration for the two-camera and three-camera tests. Each camera gets its own port; the port layout and power input are described on the product page.
Table 2. XIMEA cameras tested with the Jetson Orin NX
| Camera | Series | Sensor | Resolution | Interface | Used in tests |
| MX245CG-SY-X4G3-FF | xiX-Xtreme | Sony IMX530, color | 24.5 MP | PCIe Gen 3 x4 | 1 camera, Full HD and 4K |
| MX081CG-SY-PG4-X2G2-HDR | xiX | Sony IMX536, color, HDR | 8.1 MP | PCIe Gen 2 x2 | 2 cameras, 8.1 MP; 3 cameras, Full HD |
| MX031CG-SY-X2G2 | xiX | Sony IMX252, color | 3.1 MP | PCIe Gen 2 x2 | 3 cameras, Full HD |
The xiX family offers many more sensors in the same compact PCIe design, so resolution and speed can be matched to your application – see the xiX product page for all models.
Transmitting side:
farsight sample application
Processing pipeline
Receiving side:
Receiving application based on the IFF SDK imagebroker sample with the image processing pipeline:
Glass-to-glass latency is the time delay between the moment light hits the camera sensor and the moment the camera image is displayed on the viewer's screen. It was measured with a LatencyKit from Science Mosaic and its LatencyMeter application, over a 60 s measurement window. A blinking LED is placed in front of the cameras, and light sensors are attached to the laptop screen where the streams are shown; the application correlates both signals. In the multi-camera configurations all streams were measured simultaneously.
Picture 2. LatencyKit light sensors on the receiving laptop screen, one per camera stream
The following tests were done, each with a 1 ms exposure time, 12-bit input and 16-bit processing:
For each test the image processing time on the Orin NX, the combined camera and Orin power consumption and the glass-to-glass (G2G) latency were measured.
Table 3. NVIDIA Jetson Orin NX with XIMEA xiX cameras
| Parameter | One camera MX245CG | One camera MX245CG | Two cameras 2× MX081CG | Three cameras MX031CG + 2× MX081CG |
| Camera interface | PCIe X4G3 | PCIe X4G3 | 2× PCIe X2G2 | 3× PCIe X2G2 |
| Image size | Full HD 1920 x 1080 | 4K 3840 x 2160 | 8.1 MP 2848 x 2848 | Full HD 1920 x 1080 |
| Input image bit depth | 12-bit | 12-bit | 12-bit | 12-bit |
| Image processing bit depth | 16-bit | 16-bit | 16-bit | 16-bit |
| Exposure time | 1 ms | 1 ms | 1 ms | 1 ms |
| Frame rate (for each camera) | 390 fps | 109 fps | 55 fps | 120 fps |
| Encoder bitrate | 20 Mbps | 10 Mbps | 2× 10 Mbps | 3× 10 Mbps |
| Jetson processing time | 9 ms | 24 ms | 35 ms | 13 ms |
| Jetson power mode | MAXN Super | MAXN Super | MAXN Super | MAXN Super |
| Camera + Jetson power budget | 41 W | 42 W | 37 W | 38 W |
| Glass-to-glass latency | 28 ms | 50 ms | 69 ms | 43 ms (average) |
One MX245CG-SY-X4G3-FF camera on the PCIe Gen 3 x4 port.
Picture 3. One MX245CG camera, Full HD at 390 fps – G2G latency 28 ms
Picture 4. One MX245CG camera, 4K UHD at 109 fps – G2G latency 50 ms
Two MX081CG-SY-PG4-X2G2-HDR cameras, each on its own PCIe Gen 2 x2 port, streamed as two independent 10 Mbps H.265 streams.
8.1 MP, 2848 x 2848 images at 55 fps per camera
Picture 5. Two MX081CG cameras, 8.1 MP at 55 fps each – signal latency of both streams, G2G latency 69 ms
One MX031CG-SY-X2G2 and two MX081CG-SY-PG4-X2G2-HDR cameras, each on its own PCIe Gen 2 x2 port, streamed as three independent 10 Mbps H.265 streams. All three cameras ran at a 1920 x 1080 image size, smaller than the full sensor resolutions (2064 x 1544 and 2848 x 2848).
Picture 6. Three cameras, Full HD at 120 fps each – G2G latency 44 ms, 44 ms and 41 ms
Planning an embedded vision system with Jetson Orin? Our team will gladly help you choose the right cameras, carrier board and software. Don't hesitate to get in touch with us if you need more details about the system configuration and test results.
The tests were carried out by MRTech.
See also: NVIDIA Jetson AGX Orin benchmarks · Jetson Xavier NX benchmarks · XEC-NX-3P-X2G3 carrier board