Features

  • NVIDIA Tesla P100 GPU

    A Single card with single precision can output 9.3TFLOPS, while with double precision the card can output 4.7TFLOPS. It is equipped with 16GB HBM2 video memory, and a bandwidth of 732GB/S. The video memory interface is two times higher, and the bit width is eight times higher than the previous generation GPU.

  • NVIDIA Tesla P40 GPU

    It provides a strong single-precision floating-point computing double precision and integer computing capacity. Single peak computing capacity breaks 96 TFLOPS single-precision floating-point computing and 376 TOPS integer computing. It is equipped with 16GB HBM2 video memory, a bandwidth of 732GB/S. The video memory interface is two times higher, and the bit width is eight times higher than the previous gemeration GPU.

  • High-performance Storage

    We include the P All Series. SSD local drive deployed NVMe can provide several times more IOPS and bandwidth capacity than ordinary SSD. For large data, local NVMe SSD has extremely low access delay and high storage bandwidth, which can improve comprehensiveness.

  • High-performance Network

    GCC provides a maximum network bandwidth of 10Gb/s, and a bare-metal instance is equipped with a 100GB IB network. This combination meets the needs of the computing cluster for data transmission.

  • GPU Direct

    GCC can implement direct communication among GPU through GPU Direct technology. Data transmission efficiency can be improved by five times with NvLink technology. Data transmission thus could benefit from high bandwidth, low delay, high data efficiency, and strong comprehensive data processing ability.

Applications

  • Artificial Intelligence
  • Scientific Computing
  • Media Processing

Description

The P100GPU contains thousands of computing units and shows substantial advantages in parallel computing. For optimized deep learning, massive computing can be completed in a short time.

Advantages

• GPU Direct seamlessly supports the transmission of big data among networks.

• 100GB IB network : Support GPU Direct over RDMA, 100G ultra-high bandwidth, 2us ultra-low latency.

•  Built-in acceleration framework : It can focus on the core tasks by quick deployment and fast instance distribution.

Artificial Intelligence

Description

In Scientific Computing,it requires powerful double-precision computing capacity. During the simulation, a large amount of temporary data will be generated while consuming massive computing resources, which also has an extremely high requirement for memory bandwidth and time delay.

Advantages

• NVMe SSD : Up to 680,000 IOPS, it can avoid storage bottlenecks and improve overall performance.

• Double-precision computing:Provide 100 times more double-precision computing power than CPU.

• Seamless migration:Support various scientific computing software.

Scientific Computing

Description

powerful GPU gives you incredibly capable graphics processing abilities,especially in graphic-intensive applications like QHD video transcoding,encryption,and real-time conference.

Advantages

NVidia Tesla P40 has independent video codec hardware unit, which supports full frame content processing. A single physical GPU can support 20~40 channels (reference value depending on different parameters) H.264 HD video real-time codec.

Media Processing

Artificial Intelligence

Artificial Intelligence

Description

The P100GPU contains thousands of computing units and shows substantial advantages in parallel computing. For optimized deep learning, massive computing can be completed in a short time.

Advantages

• GPU Direct seamlessly supports the transmission of big data among networks.

• 100GB IB network : Support GPU Direct over RDMA, 100G ultra-high bandwidth, 2us ultra-low latency.

•  Built-in acceleration framework : It can focus on the core tasks by quick deployment and fast instance distribution.

Scientific Computing

Scientific Computing

Description

In Scientific Computing,it requires powerful double-precision computing capacity. During the simulation, a large amount of temporary data will be generated while consuming massive computing resources, which also has an extremely high requirement for memory bandwidth and time delay.

Advantages

• NVMe SSD : Up to 680,000 IOPS, it can avoid storage bottlenecks and improve overall performance.

• Double-precision computing:Provide 100 times more double-precision computing power than CPU.

• Seamless migration:Support various scientific computing software.

Media Processing

Media Processing

Description

powerful GPU gives you incredibly capable graphics processing abilities,especially in graphic-intensive applications like QHD video transcoding,encryption,and real-time conference.

Advantages

NVidia Tesla P40 has independent video codec hardware unit, which supports full frame content processing. A single physical GPU can support 20~40 channels (reference value depending on different parameters) H.264 HD video real-time codec.

Advantages

Cost-effective

Cost-effective

GCC synchronizes the latest GPU technology in the field, and seamlessly switches the latest GPU hardware. GCC offers different billing methods including on-demand, and periodical, rent-and-use, and elastic expansion.

Safety Protection

Safety Protection

The resources among different users are completely isolated to ensure data security. Also, a network monitoring service provides network security. GCC is seamlessly connected with cloud security and shares the same basic cloud security and high defense services of cloud servers.

Cost-saving

Cost-saving

You can reduce infrastructure investment by avoiding frequent hardware updates.

Simple

Simple

You can focus on the core tasks through quick access to various supercomputing applications and deep learning frameworks.

High Performance

High Performance

GCC with extreme performance breaks through traditional GPU, which has high parallel, high throughput, low latency. Performance in scientific computing is fifty times faster than traditional architecture.

24/7 Support
Contact sales
Chat Now