算力服务

云算力支持17台高性能服务器，136张GPU卡云算力服务；GPU计算性能达到单卡NVIDIA A100，显存80G及以上；机器学习平台：平台面向机器学习应用开发者，提供开发机、自定义任务、实验管理等丰富的建模调试工具以及多框架高性能推理服务的企业级云原生机器学习平台；高性能计算GPU 云服务器：提供 GPU 算力的弹性计算服务，具有高效稳定的计算能力，适用于生成式AI、自动驾驶、图像处理、科学计算等多种应用场景；弹性块存储：提供的高可用、高可靠、高性能的块存储设备，可以作为云服务器和弹性容器服务的可扩展硬盘使用；对象存储：提供的海量、安全、低成本、易用、高可靠、高可用的分布式云存储服务。通过网络，您可以在任何应用、任何时间、任何地点管理和访问对象存储上的数据；文件存储：是一种高吞吐、低延时、可扩展的并行文件存储服务，满足高性能计算场景高吞吐低延时的数据读写需求，可广泛应用于HPC高性能计算、AI人工智能训练/推理等场景；云服务器：是一种安全稳定、可弹性伸缩的云计算服务，根据业务需要实时启用和增减计算资源，有效降低IT维护成本，助力核心业务增长。

The cloud computing infrastructure supports 17 high-performance servers and 136 GPU cards for cloud computing services; each GPU has the computing performance of an NVIDIA A100, with 80GB or higher video memory.
Machine Learning Platform: An enterprise-grade cloud-native machine learning platform tailored for machine learning application developers, which provides rich modeling and debugging tools including development environments, custom tasks, experiment management, as well as high-performance inference services supporting multiple frameworks.
High-Performance Computing GPU Cloud Servers: Elastic computing services that provide GPU computing power, featuring efficient and stable computing capabilities, suitable for various application scenarios such as generative AI, autonomous driving, image processing, and scientific computing.
Elastic Block Storage: High-availability, high-reliability, and high-performance block storage devices that can be used as scalable hard disks for cloud servers and elastic container services.
Object Storage: A distributed cloud storage service featuring massive capacity, security, low cost, ease of use, high reliability, and high availability. Through the network, users can manage and access data stored on the object storage from any application, at any time, and from any location.
File Storage: A high-throughput, low-latency, scalable parallel file storage service that meets the high-throughput and low-latency data read/write requirements of high-performance computing scenarios, and can be widely used in HPC high-performance computing, AI training/inference and other scenarios.
Cloud Servers: A secure, stable, and elastically scalable cloud computing service that enables real-time activation and scaling of computing resources according to business needs, effectively reducing IT maintenance costs and supporting the growth of core businesses.

搜集汇总

背景与挑战

背景概述

该数据集描述了一个提供GPU云算力服务的平台，包含17台高性能服务器和136张A100显卡，配套机器学习开发工具及多种存储服务，适用于AI训练、科学计算等高性能计算场景。