◆ Support simultaneous processing of 16 channels of 1080P@30fps H.264/H.265 real-time stream or storage stream.
◆ Support 16Tops deep learning operations, using the combination of front-end image preprocessing + back-end algorithm operations, and support mainstream deep learning frameworks such as Caffe, Tensorflow and Mxnet.
◆ Support face detection, tracking and recognition.
◆ Support dual Gigabit Ethernet ports, expandable 4G/WIFI wireless network mode.
◆ Fully embedded structure and highly integrated. Support desktop placement or wall-mounted installation, which make the installation and maintenance become very convenient.
◆ Support simultaneous processing of 16 channels of 1080P@30fps H.264/H.265 real-time stream or storage stream.
◆ Support 16Tops deep learning operations, using the combination of front-end image preprocessing + back-end algorithm operations, and support mainstream deep learning frameworks such as Caffe, Tensorflow and Mxnet.
◆ Support face detection, tracking and recognition.
◆ Support dual Gigabit Ethernet ports, expandable 4G/WIFI wireless network mode.
◆ Fully embedded structure and highly integrated. Support desktop placement or wall-mounted installation, which make the installation and maintenance become very convenient.
IVP02D AI computing box is an intelligent video analysis terminal supporting deep learning neural network algorithm. Integrates high-performance 16Tops AI processing module, IVP02D is capable of 16x 1080P HD video encoding and decoding. IVP02D is featured in lightweight, high-performance and low power consumption. It has been widely used in the fields of human face recognition, smart security and behavior analysis.
Item | Type | Model | Description | |
Core Specification | Processor | MLU220 | CPU | Quad-Core ARM Cortex-A55 |
NPU | Support INT16, INT8. 16Tops computing power | |||
RAM | LPDDR4 | 8GB (Customizable) | ||
FLASH | eMMC | 32GB (Customizable) | ||
Flash | 16MB Nor Flash | |||
Function | Video Encoding | 8x | 8x 1080@30fps Support H.264/H.265/VP8/VP9 | |
Video Decoding | 16x | 16x 1080@30fps Support H.264/H.265/VP8/VP9 | ||
Image Decoding | 1s | 820 fps@FHD Support JPEG decode Support 4:0:0/4:2:2/4:2:0/4:4:0/4:4:4 format | ||
Image Encoding | 1s | 800 fps@FHD Support JPEG decode Support 4:0:0/4:2:2/4:2:0/4:4:0/4:4:4 format | ||
GMAC | 2x | Support 10/100/1000Mbps Support RGMII | ||
RS485 | 1x | 3-pin connector | ||
RS232 | 1x | 3-pin connector | ||
GPIO | 2x | 6-pin connector | ||
M.SATA (Optional) | 1x | |||
USB3.0 | 2x | |||
Debug Port | 1x | CPU_UART; Type C interface | ||
WIFI | 1x | BL-M8821CS1 | ||
TF Card Slot | 1x | |||
HDMI Output | 1x | 1080P resolution | ||
LED Indicator | 4x | Power indicator | ||
M2 | 1x | Support 5G module (Optional) | ||
Memory Interface Frequency | Up to 3733Mb/s | |||
Cooling | Passive cooling | |||
Power | Power Input | DC 12V |
Base Address | BAR0: 16MB non-prefetchable BAR2: 32MB preftchable BAR4: 32MB prefetchable |
ECC Protection | Not supported |
SMBus (8bit Address) | 0x8E(write) 0x8F(read) |
Decoding Capability | 16x 1080@30fps |
Encoding Capability | 8x 1080@30fps |
OS | Linux |
IVP02D fully supports various mainstream programming frameworks (such as TensorFlow, Caffe, Caffe2, MXNet, ONNX, etc.). Users can easily develop and deploy deep learning applications on AIM16T for the above-mentioned programming framework. At the same time, NeuWare provides a complete runtime system and driver software to facilitate rapid system integration.
IVP02D also provides a series of tools including application development, function debugging, performance tuning, etc. Among them, application development tools include machine learning libraries, runtime libraries, compilers, model retraining tools, and SDKs for specific fields (such as video analysis fields); functional debugging tools can meet different levels of debugging needs such as programming frameworks and function libraries; performance Tuning tools include performance analysis tools and system monitoring tools.
enquiry form:
enquiry content:
You have no items to require
enquiry form:
enquiry content:
You have no items to require