Deep Learning Cluster Server

QY-S1X-384 is equipped with SOPHON fourth-generation TPU processor BM1684X, which is an edge computing server for deep learning inference applications. It is widely adapted to third-party algorithms to implement applications such as face recognition, video structuring, behavior analysis, and status monitoring, in order to realize the deep learning empowerment for smart cities, smart transportation, smart energy, smart finance, smart telecommunications, and smart industries.

Specification and Characteristic

Product Model
QY-S1X-384
Application Scenarios

Smart city, Smart transportation, Smart energy, Smart finance, Smart telecommunications, and Smart industries

Deep Learning Processor
SOPHON BM1684X
Deep Learning Performance
FP32/FP16/BF16/INT8 supported, 384-channel HD video decoding and analysis
Master Processor
8-core ARM A53, basic frequency up to 2.3GHz
Memory
192GB 4000MHz LPDDR4x
Video Decoding Capability
H.265/H.264 1080p@9600fps; Resolution: 8192 * 8192 / 8K / 4K / 1080P / 720P / D1 / CIF
Video Encoding Capability
H.265/H.264 1080p@3600fps; Resolution: 4K / 1080P / 720P / D1 / CIF
Video Transcoding Capability
384 channels (Maximum)
Picture Codec Capability
7200 images/second @1080P
Storage
32GB eMMC, 2.5-inch SATA disk port available(inside)
Network Interface
10/100/1000Mbps *3
USB Interface
2 x USB3.0
Other Interface
Micro USB*1 / Micro SD *1
Power
2 x 500W, support 1+1 redundancy mode
Size
D(362mm) *W(442mm)*H(44.5mm)
Operating Temperature
+5℃ ~ +45℃
Operating System
Ubuntu
Deep Learning Framework
TensorFlow / Caffe / PyTorch / Paddle / ONNX / MXNet / DarkNet

Partner Introduction

Beijing Qiangyun Innovation Technology Co., Ltd Official Website

     Beijing Qiangyun Innovation Technology Co., Ltd (hereinafter referred to as" Qiangyun “or" the company "). was established in 2008, headquartered in Beijing Big Data Intelligent Industrial Park, and has branches in Shenzhen and Changsha. Qiangyun follows the trend of High-technology, focusing on  deep learning products manufacturing and scenario-based solutions. It has excellent products and high reputation in the field of artificial intelligence hardware, mainly providing customers with integrated solutions including hardware, software and algorithm. Its strong R & D team and self-built factories enable the company to mass produce the intelligent devices.