SOPHON Intelligent Computing Server

Datacenter

SOPHON Intelligent Computing Server

SOPHON SG6-10-B22 intelligent computing server can be equipped with 30 SOPHGO third-generation TPU processor BM1684 and its INT8 performance up to 528TOPS. It supports 28800fps 1080P HD video hardware decoding and 1500fps 1080P HD video hardware encoding. FOR PRICE INQUIRY CONTACT US

$0.10

Add to Cart

Empowering Technology

Incredible Computing Efficiency

Strong Computing Power

Provide up to 528TOPS of INT8

computing power or 66TFLOPS of FP32 computing power

Ultra High Performance

960 channels of full HD video decoding

60 channels of full HD video hardware encoding

Video and image decoding resolution up to over 8K

Rapid Transplantation

Support mainstream deep learning frameworks such as

Caffe/TensorFlow/PyTorch/ MXNet/Paddle Lite

Advanced Scalability and Compatibility

Enhancing System Performance

Flexible Extension

Support 8*2.5” front hot-swappable hard disk slots

(2 of which support NVME)

10*PCIE3.0 extendible slots

8*DDR4 memory slots, supporting the maximum capacity of 1TB

High compatibility

AMD EPYC X86 processor platform

Support RESTful/IPMI management interface

Wide Application and Scenarios

Smart Adaptability

Adapt to multiple intelligent algorithms and support intelligent analysis applications such as face recognition, vehicle recognition, item recognition, video structurization and behavior analysis.

Easy-to-Use, Convenient and Efficient

SOPHON SDK one-stop toolkit provides a series of software tools including the underlying driver environment, compiler and inference deployment tool. The easy-to-use and convenient toolkit covers the model optimization, efficient runtime support and other capabilities required for neural network inference. It provides easy-to-use and efficient full-stack solutions for the development and deployment of deep learning applications. SOPHON SDK minimizes the development cycle and cost of algorithms and software. Users can quickly deploy deep learning algorithms on various deep learning hardware products of SOPHGO to facilitate intelligent applications.

Support mainstream programming framework

Performance Parameter

Table 1

Aspect	Details
Deep Learning Performance	Supports FP32/INT8, up to 480 HD video channels
Video Decoding	H.264: 1080P @28800fps / H.265: 1080P @28800fps
Video Encoding Resolution	8192 * 8192 / 8K / 4K / 1080P / 720P / D1 / CIF
Video Transcoding	H.264: 1080P @1500fps / H.265: 1080P @1500fps
Picture Decoding	4K / 1080P / 720P / D1 / CIF
Picture Max Resolution	Max 540 channels (1080P to CIF)
Video Memory	24000 JPEG/sec @1080P

System Specifications

Table 2

Aspect	Details
Coprocessor	ARM A53 240 core @2.3GHz
Processor	1 AMD EPYC 7002 series
RAM	8 * DDR4 DIMM 3200MHz, up to 1TB
Storage	6 2.5" SATA/SAS HDDs / 2 2.5" U.2 NVMe/SATA SSDs
Interfaces	2 SFP+ 10G LAN / 2 x USB3.0 / 1 x VGA / 1 x MGMT
Power Supply	Two 2200W 80 PLUS platinum PSUs (1+1 redundancy)
Heat Dissipation	6 fans