Series:A800 series

  Product Description

  The A800 NVLink NVIDIA computing card is a computing accelerator card used for high-performance computing and deep learning tasks.It has excellent computing performance and energy efficiency,and supports NVLink technology to achieve high-speed data transmission and communication between GPUs.

  feature

  Support NVLink technology

  Capable of high-speed data transmission and communication between GPUs

  The A800 NVLink NVIDIA computing card is designed specifically for users who require high-performance computing and deep learning capabilities.It provides powerful computing performance and excellent energy efficiency,suitable for various scientific research,artificial intelligence,and data analysis tasks

  NVIDIA®NVLink®Technology can provide 1.5 times higher bandwidth and enhanced scalability for multi GPU system configurations than before.A single NVIDIA H100 Tensor Core GPU supports up to 18 NVLink connections,with a total bandwidth of 900 GB/s,which is 7 times the bandwidth of PCIe 5.0.

  NVIDIA DGX?H100 and other servers can use this technology to improve scalability and achieve ultra fast deep learning training.

  NVIDIA NVSwitch?Building advanced communication capabilities based on NVLink can provide higher bandwidth and lower latency for computationally intensive workloads.To support high-speed set operations,each NVSwitch has 64 NVLink ports and is equipped with NVIDIA SHARP?Engine,which can be used for network reduction and multicast acceleration.

  With NVSwitch,NVLink connections can be extended between nodes to create seamless,high-bandwidth,multi node GPU clusters,effectively forming data center sized GPUs.By adding a second layer NVSwitch outside the server,the NVLink network can connect up to 256 GPUs and provide an astonishing many to many bandwidth of 57.6 TB/s,enabling rapid completion of large-scale AI tasks.

  Achieve exceptional performance through full connectivity

  NVSwitch is the first node switching architecture that can support 8 to 16 fully interconnected GPUs in a single server node.The third-generation NVSwitch can interconnect each pair of GPUs at an astonishing speed of 900GB/s.It also supports complete many-to-many communication.These GPUs can be used as a single high-performance accelerator with deep learning computing performance of up to 15 petaFLOPS.

  Powerful AI and HPC platforms

  NVLink and NVSwitch are essential foundational modules for the entire NVIDIA data center solution,which includes software packages from NVIDIA AI Enterprise and NVIDIA NGC?The hardware,network,software,libraries,and optimized AI models and applications of the directory.As a powerful end-to-end AI and high-performance computing platform,it enables researchers to obtain faster results and deploy solutions into production environments,achieving excellent acceleration on various scales.

  application area

  The A800 NVLink NVIDIA computing card is an ideal choice for high-performance computing,artificial intelligence,deep learning,and data analysis applications.

  The United States has introduced new regulations on semiconductor export restrictions to China,including restrictions on the export of high-performance computing chips to Chinese Mainland.And the performance indicators of NVIDIA's A100 chip are used as limiting standards.A regulated high-performance computing chip is one that meets both of the following conditions:

  (1)The I/O bandwidth transmission rate of the chip is greater than or equal to 600 Gbyte/s;

  (2)The sum of the bit length of each operation multiplied by the computing power calculated by TOPS in the"original computing unit of the digital processing unit"is greater than or equal to 4800 TOPS.This also makes it impossible for NVIDIA A100/H100 series and AMD MI200/300 series AI chips to be exported to China.