APX-3323 GPU
CL, Base/ Medium/Full GPU installed Frame Grabber

  • Intelligent Video Processing Board by GPU+CPU (ARM)
  • NVIDIA Corporation Jetson TX 1
  • Connect up to 1 Camera Link BASE / Medium / Full Configuration camera
  • M.2SSD and micro SD can be installed as storage
  • CUDA developed as a general purpose GPU product is available
  • Many free CUDA libraries are available
  • Can transfer image directly to GPU memory
  • PCI Express I / F supports PCI Express GEN 2 (5 Gbps) 4 lanes
Model APX-3323GPU-TX1
Image processing processor NVIDIA Jetson TX1 SOM (System On Module)
Tegra X1 installed
(64 bit ARM Cortex – A 57 CPU, 256 Core GPU
(Maxwell architecture))
Panel I / F Image input I / F Camera Link Base / Medium / Full Configuration×1 ch
(SDR 26 pin connector × 2, PoCL compatible)
USB3.0 1ch
LAN 1ch
MEMORY DDR2-SDRAM 256MB (Camera Link Controller Work Memory)
Onboard I / F I / O APX-3302 compatible external trigger input /
general input / encoder input
COM 1ch (BOX HEADER 10pin)
USB2.0 1ch
micro SD 1ch
mini DisplayPort 1ch
Event Image input start, DMA end, GPIN etc.
System bus PCI Express2.0 × 4 (Gen2) 5.0GT / S
Power supply + 12 V ± 8%
* Using external + 12 V connector (for PoCL)
Power + 12V: 2A Max ※ excluding PoCL
Environment Temperature: 0 ° C to 50 ° C
Humidity: 35% to 80% (no condensation)
Dimension (substrate) 254mm × 111.15mm (2slot)
Compliance RoHS
Software overview GPU
Jetson
Development Pack		 JetPack 2.3.1
OS Ubuntu 16.04 LTS 64bit, Kernel Ver 3.10.96
CUDA Ver 8.0 (8.0.34)
Accessory tool VisionWorks 1.5, Tegra System Profiler,
Tegra Graphics Debugger, OpenCV4Tegra 2.4.13,
Multimedia API v24.2.1, cuDNN v5.1, Tensor RT
SDK-AcapLib2-TX1 Linux Capture SDK
※Pre-installed
Note) Not compatible with Windows
Notes)
– The specifications and appearance stated are subject to change without notice.
– Power consumption does not include inrush current.
– In order to use the product correctly, be sure to read the instruction manual before using the product. Also, use the product under the conditions applicable to your product warranty

Since Linux (Ubuntu), Linux SDK, and CUDA development environment (CUDA Driver) are already installed on the APX – 3323 GPU, you can develop applications onboard that controls the grabber and process the GPU.

APX-3323GPU and Deep Learning can be applied for the external examination.