Alibaba Cloud MaxCompute - Bringing Data to Life

简介:

NW_001

The value of computing is not just to calculate, but to discover hidden trends in data. From the Mayan calendar to the circumference ratio. From Newton's law of universal gravitation to binary code, and even from creative ideas to concrete plans. Data is everywhere, and Alibaba Cloud's MaxCompute is a technological-driven product that uses supercomputing to bring that data to life.

Formerly known as ODPS, Alibaba Cloud's MaxCompute is Alibaba Cloud’s primary computing platform. From ODPS to MaxCompute, it stores 99% and handles 95% of Alibaba Group's storage and big data computing. Moreover, it has integrated such technologies as the public cloud and Apsara Stack. Through continuous technological innovation, MaxCompute’s computing capability never stops improving and has become a powerful engine for data transformation that spans all industries and sectors of society. Helping enterprises, large and small, to turn valuable data into actionable insights, plans, and decisions. MaxCompute has an excellent reputation for using its supercomputing for both social benefits and unparalleled business results.

Alibaba Cloud’s MaxCompute offers efficient and fully collocated PB level inventory solution. While MaxCompute might look complicated from the outside, in actuality it has penetrated every aspect of society. For example, it supports MaxCompute powers an array of services, applications, and businesses such as your local weather channel forecasting, healthcare systems, live streaming and even your daily transportation.

One market that shows MaxCompute’s true potential is the global bike sharing industry. In 2017, bike-sharing exploded onto the world's market, and today the number of bikes offered by bike-sharing companies like Ofo is skyrocketing. Offline, Ofo and others are using MaxCompute’s data processing power for bike scheduling, location data, and network stability and have witnessed a 50% increase in operational efficiency. Online, bike-sharing corporations can quickly deal with millions of daily requests, and those same companies overall operational productivity has increased by 76%. MaxCompute continues to offer more than four billion bike-sharing users from all over the world a ride on one of ten million bikes.

MaxCompute can also be used to enhance target advertising. In one example a weather App processed millions of weather updates and billions of requests for promotions. Businesses can rely on MaxCompute to handle this type of volume; in fact, overall business efficiency can be improved by more than five-times by using MaxCompute. Functions and stability can be improved, and operational costs are cut by 70 percent on average.

When it comes to the insurance industry, MaxCompute helps enterprises accurately analyze user profiles and offer hundreds of millions of users billions in insurance policies. While MaxCompute improves business efficiency, it also provides hundreds of millions of users with health and safety.

Apart from people's livelihood, MaxCompute provides computing capabilities to high-tech gene sequencing research laboratories. Traditional computing cannot compute such massive data stores. However, MaxCompute can finish this task within several hours. Furthermore, MaxCompute is economical while still leading the computing engine field of gene sequencing.

While only a big data computing engine, MaxCompute can bring significant social and business benefits. MaxCompute seeks not only to bring the value of computing to companies but to society at large. Because supporting each successful business or social benefit are computing capabilities like MaxCompute that will keep improving and innovating.

In 2017, MaxCompute has achieved several breakthrough technologies.

• MaxCompute has adapted TPC benchmark and increased data size to 100 Terabytes (TB). It has accomplished the benchmark testing of BigBench based on public cloud for the first time in the world, and its performance broke 8200 QPM. The computing power of MaxCompute is not only the leader in China but also the rest of the world.

• Regarding off-line computing, MaxCompute integrates Lightning Compute into interactive computing modes. Combining off-line data with real-time data, MaxCompute cut the average request response time to just milliseconds.

• The creation and development of AliOrc upgraded MaxCompute's architecture and computing power. It ultimately produced a highly available open platform that is self-adaptable which can analyze NewSQL databases, structured and unstructured data.

• To increase resource utilization, MaxCompute uses a separation architecture and a hybrid deployment of storage and computing. During 2017’s Double 11 Shopping Festival, the online peak data of E-commerce with a hybrid deployment was 6.5W deals/s (20% transactional data), and the peak data of Ant Financial with a hybrid deployment was 5W deals/s.

• Datahub architecture has been upgraded and successfully supports video and media. On November 11, 2017, DataHub hosted the data collection, consumption, and delivery on the PB level. There was never a delay that lasted over two seconds. On Double 11 Shopping Festival, the technology was unveiled, signifying the first time data recovery was achieved across data centers with such consistent results. In case of a system breakdown, MaxCompute can reset in minutes and data is recovered instantaneously.

• To provide high-quality services of big data computing to the global market and clients, in 2017, MaxCompute accelerated its deployments around the world and already provides services for over ten regions across the globe. In the first half year of 2018, MaxCompute will provide services to India, Indonesia as well as other overseas countries.

Alibaba Cloud MaxCompute pioneers the goal of bringing cloud computing and big data to the business sectors. It is vital to realize the power of data and analytics in a business setting, and MaxCompute shows its full value. To learn more about MaxCompute, please visit our website for more information.

相关实践学习
简单用户画像分析
本场景主要介绍基于海量日志数据进行简单用户画像分析为背景,如何通过使用DataWorks完成数据采集 、加工数据、配置数据质量监控和数据可视化展现等任务。
SaaS 模式云数据仓库必修课
本课程由阿里云开发者社区和阿里云大数据团队共同出品,是SaaS模式云原生数据仓库领导者MaxCompute核心课程。本课程由阿里云资深产品和技术专家们从概念到方法,从场景到实践,体系化的将阿里巴巴飞天大数据平台10多年的经过验证的方法与实践深入浅出的讲给开发者们。帮助大数据开发者快速了解并掌握SaaS模式的云原生的数据仓库,助力开发者学习了解先进的技术栈,并能在实际业务中敏捷的进行大数据分析,赋能企业业务。 通过本课程可以了解SaaS模式云原生数据仓库领导者MaxCompute核心功能及典型适用场景,可应用MaxCompute实现数仓搭建,快速进行大数据分析。适合大数据工程师、大数据分析师 大量数据需要处理、存储和管理,需要搭建数据仓库?学它! 没有足够人员和经验来运维大数据平台,不想自建IDC买机器,需要免运维的大数据平台?会SQL就等于会大数据?学它! 想知道大数据用得对不对,想用更少的钱得到持续演进的数仓能力?获得极致弹性的计算资源和更好的性能,以及持续保护数据安全的生产环境?学它! 想要获得灵活的分析能力,快速洞察数据规律特征?想要兼得数据湖的灵活性与数据仓库的成长性?学它! 出品人:阿里云大数据产品及研发团队专家 产品 MaxCompute 官网 https://www.aliyun.com/product/odps 
目录
相关文章
|
5月前
|
存储 分布式计算 运维
【2023云栖】刘一鸣:Data+AI时代大数据平台建设的思考与发布
本文根据2023云栖大会演讲实录整理而成,演讲信息如下: 演讲人:刘一鸣 | 阿里云自研大数据产品负责人 演讲主题:Data+AI时代大数据平台应该如何建设
101381 7
|
10月前
|
存储 SQL 机器学习/深度学习
MaxCompute(原名ODPS,全称Open Data Processing Service)
MaxCompute(原名ODPS,全称Open Data Processing Service)是阿里云开发的一种云原生数据处理和分析服务。它提供了强大的数据计算和处理能力,支持海量数据的存储、计算、分析和挖掘,并且具有高可靠、高性能、高可扩展、高安全等优势,适用于各种数据处理和分析场景。
573 0
|
存储 运维 Oracle
【大数据开发运维解决方案】Oracle Data Redaction数据加密测试
最近有个做Java开发的网友问我,怎么在Oracle进行数据加密呢?我给他推荐了Data Redaction。Oracle Database 12c中加入了Data Redaction这个新的安全特性。当然在11g的Database Advanced Security Administrator’s Guide官方文档中就介绍了。
【大数据开发运维解决方案】Oracle Data Redaction数据加密测试
|
大数据
阿里云大数据ACP(二)数据集成 Data Integration 2
阿里云大数据ACP(二)数据集成 Data Integration 2
131 0
阿里云大数据ACP(二)数据集成 Data Integration 2
|
DataWorks 安全 数据可视化
阿里云大数据ACP(二)数据集成 Data Integration 1
阿里云大数据ACP(二)数据集成 Data Integration 1
394 0
阿里云大数据ACP(二)数据集成 Data Integration 1
|
消息中间件 SQL 分布式计算
IDEA 中使用 Big Data Tools 连接大数据组件
简介 Big Data Tools 插件可用于 Intellij Idea 2019.2 及以后的版本。它提供了使用 Zeppelin,AWS S3,Spark,Google Cloud Storage,Minio,Linode,数字开放空间,Microsoft Azure 和 Hadoop 分布式文件系统(HDFS)来监视和处理数据的特定功能。 下面来看一下 Big Data Tools 的安装和使用,主要会配置 Flink,Kafka 和 HDFS。
IDEA 中使用 Big Data Tools 连接大数据组件
|
存储 数据采集 人工智能
初始大数据(Big Data)开发
大数据(big data),或称巨量资料,指的是所涉及的资料量规模巨大到无法透过目前主流软件工具,在合理时间内达到撷取、管理、处理、并整理成为帮助企业经营决策更积极目的的资讯。主要解决的是对海量数据的存储以及海量数据的计算分析问题
初始大数据(Big Data)开发
|
机器学习/深度学习 算法 Java
大数据data开发有哪些好的辅助工具?
作为一个程序员开发工具好比是人的手和脚,只有把这些开发工具用好,才能做好一个产品的需求。大多使用SQL数据库存储/检索数据,如今很多情况下,它都不再能满足我们的需求。下面小编就介绍一些大数据data开发常用的辅助工具。
|
机器学习/深度学习 存储 人工智能
被神话的大数据——从大数据(big data)到深度数据(deep data)思维转变
自从阿法狗战胜人类顶级棋手之后,深度学习、人工智能变得再一次火热起来,但有一个基本的误解是更大的数据会产生更好的机器学习结果。然而,更大的数据池/仓库并不一定有助于模型学习到更深刻的见解。正确的答案是?
2052 0
|
存储 分布式计算 大数据
阿里云大数据ACP认证知识点梳理9——产品特点(DATA WORKS)
DATA WORKS(原DATA IDE) 产品特点及重点注意事项
3082 1

相关产品

  • 云原生大数据计算服务 MaxCompute