Elastic GPU Service provides world-leading coverage, superior computing capacity, excellent network performance, and flexible purchase methods.

  • World-leading coverage

    With large-scale deployment in 17 regions, Elastic GPU Service provides world-leading coverage. Based on delivery methods such as auto supply and auto provisioning, Elastic GPU Service is able to handle emergent needs of your business.

  • Superior computing capacity

    Elastic GPU Service is equipped with industry-leading GPU processors. Combined with a high-performance CPU platform, a single instance with GPU capabilities can provide mixed precision computing performance of up to 1,000 TFLOPS.

  • Excellent network performance

    VPCs of instances with GPU capabilities support up to 4,500 Kpps and 32 Gbit/s internal bandwidth. In addition, Super Computing Cluster (SCC) products provide an additional RDMA network of up to 50 Gbit/s between nodes to meet the low-latency and high-bandwidth requirements for data transmission between nodes.

  • Flexible purchase methods
    Elastic GPU Service supports a variety of billing methods including subscription, pay-as-you-go, preemptible instances, and storage capacity units (SCUs). You can select the billing method that best suits your needs to avoid inefficient use of resources.
    Note Reserved instances cannot be used for instances with GPU capabilities.