The SOPHON AI computing box SE5 is a high-performance, low-power edge computing product. It is equipped with the third-generation TPU chip BM1684 independently developed by Bitmain. The INT8 computing power is up to 17.6TOPS. It can support up to 38 channels of 1080P HD video hardware decoding and 2 channels encoding.
Support up to 17.6T INT8 peak computing power or 2.2T FP32 high-precision computing power
Support up to 16 channels 1080P HD video full-process processing, support 38 channels 1080P HD video hardware decoding and 2 channels encoding
Support Caffe / TensorFlow / PyTorch / MXNet / Paddle Lite and other mainstream deep learning frameworks
Support fanless cooling, wide temperature range from -20℃ to + 60℃
Support various algorithms such as person / vehicle / non-vechicle / object recognition, video structuring, trajectory behavior analysis, etc.
Support smart Park / security / industrial control / business and other fields and scenarios for flexible deployment
Support USB, HDMI, RS-485, RS-232, SATA, custom I/O and other interfaces
Support LTE wireless function (optional)
Video stream and image stream recognition, video structure analysis, trajectory analysis, etc.
Checkpoint, automatic vechicle identification, law enforcement assistance, smart parking
Face recognition security checking, staff attendance management, smart alart system
Goods recognition, smart stores application, face recongnization payment
BMNNSDK (BITMAIN Neural Network SDK) one-stop toolkit provides a series of software tools such as the underlying driver environment, compiler, inference deployment tool and so on. Easy to use and convenient, covering the model optimization, efficient runtime support and other capabilities required for the neural network inference stage, providing easy-to-use and efficient full-stack solutions for deep learning application development and deployment. BMNNSDK minimizes the development cycle and cost of algorithms and software. Users can quickly deploy deep learning algorithms on various AI hardware products of Fortune Group to facilitate intelligent applications.
chip
model
SOPHON BM1684
Computing power
INT8
Peak computing power 17.6TOPS
FP32
Peak computing power 2.2TOPS
Video / Picture codec
Video decoding capability
960fps 1080p (38 channels 1080P @ 25FPS)
Video encoding capability
50fps 1080p (2 channels 1080P @ 25FPS)
Picture codec capability
480 PCS / s 1080p
Memory and storage
Memory
12GB
eMMC
32GB
External interface
Ethernet interface
10/100 / 1000Mbps adaptive * 2
USB
USB3.1 *2
Storage
MicroSD *1
Display
HDMI *1
Phoenix terminal
RS-232 * 1 / RS-485 * 1 / Custom I/O
Mechanical
Length * width * height
188mm*148mm*44.5mm
Power supply and power consumption
Power supply
DC 12V
Typical power consumption
≤20W
Temperature and humidity
Working temperature
-20 ℃ ~ + 60 ℃ (depending on the configuration)
Humidity
10% ~ 90%, no condensation
Other functions
Optional
SATA hard disk support
LTE wireless backhaul support