从亚马逊云科技 re:Invent 2024 看数据与AI的未来

“We invent so you can reinvent”

从亚马逊云科技 re:Invent 2024 看数据与AI的未来

The Future of Data and AI: Insights from Amazon Web Services re:Invent 2024

 

引言/大会概述

Introduction/Conference Overview

在All in AI的时代下，是做一个更好的工具平台，还是做强而有力的产品，赋能自己还是赋能他人，自我创新与为他人的创新铺路。可以说是头部科技巨头的业务发展的主要长期策略逻辑。

In the era of "All in AI", technology giants face crucial choices: to build a better tool platform or powerful products, to empower themselves or others, and to focus on self-innovation or pave the way for others' innovation. These are the main long-term strategic logics for the business development of leading tech behemoths.

Open AI的出世及GPT的表现一方面的确对长期处于领先优势的亚马逊云科技无疑有一定的冲击，很多用户因为强大的GPT，选择与微软合作构建长期的AI战略，亚马逊云科技作为云计算产业的开创者一直以来在产品创新及领导，新技术的突破上作为行业的标杆，虽然在时机上有一定的滞后，但正如我们所看，真正伟大的技术和产业的发展变革中，早一天和晚一天并不会从本质上带来影响，如何准确的寻找自己的战略位置，以及如何带来价值永远是市场竞争中的主要影响因素。

The emergence of OpenAI and the performance of GPT have undoubtedly had a certain impact on Amazon Web Services (AWS), which has long held a leading position. Many users, attracted by the powerful GPT, have chosen to cooperate with Microsoft to build long-term AI strategies. As the pioneer of the cloud computing industry, AWS has always been a benchmark in product innovation, leadership, and breakthroughs in new technologies. Although it was somewhat late to the game, as we can see, in the truly great development and transformation of technologies and industries, a day or two earlier or later doesn't essentially matter. Accurately finding one's strategic position and bringing value are always the main influencing factors in market competition.

整体策略上可以看到，亚马逊云科技的战略定力，体现在

Overall, AWS's strategic focus is reflected in the following aspects:

帮助他人创新，给用户更好的工具箱

Empowering Others to Innovate and Offering Better Toolkits

与用户需求为导向：产品的优化，可以看到的是几个产品的改进是深度贴合用户需求的，比如Model Distillation模型蒸馏；Bedrock Automated Reasoning check推理检查；multi-agent collaboration多智能体协作，这三款功能性产品的发布，可以让我们切实的看到从去年以来Bedrock系列的发布，有大量的用户在使用，并且在使用中，亚马逊云科技也快速的响应其中用户的需求，将使用中的共性问题沉淀出产品及解决方案。

User-Centric: Product optimizations are deeply tailored to user needs. For example, features like Model Distillation, Bedrock Automated Reasoning check, and multi-agent collaboration have been released. The release of these three functional products shows that since the launch of the Bedrock series last year, a large number of users have been using them. Moreover, AWS quickly responds to user needs during usage and turns common problems into products and solutions.

Choice ：永远给予用户开放的自主选择权，发布的nova系列，同时也有其他家的大模型在平台里，给到用户更多更好的产品。比起一家独大的训练吵架AGI，如何给到用户自主选择的权力，同时如何启发用户去做更加丰富的开发和应用，同时将“选择权”下放，这产品背后的核心逻辑更加适应于百花齐放的科技格局。

Choice: Always giving users open and autonomous choices. The released nova series, along with other large models on the platform, provides users with more and better products. Compared to a monopolistic approach to AGI training, AWS focuses on how to give users the right to choose, inspire them to develop and apply more diversely, and decentralize the "right to choose". This core logic behind the products is more adaptable to the flourishing technology landscape.

长期的具有前瞻性的产品策略：产品之间的开放性和完整性使得所有的产品可以提前的为未来的新技术留下灵活变动的可能性,sagemaker AI 在原有的AI开发平台上融入了新的生成式AI的能力。Sagemaker在过去的成功，让众多开发者仍旧喜爱这项开发平台，在新的GenAI时代下，如何不摒弃原有的产品，而是将原有的产品可以快速融合新的技术，这对于原有的产品涉及以及新技术发展，新产品推出都有极大的技术及产品能力的考验，需要过去的产品有一定的开放性和前瞻性，这其中的特点也可以在亚马逊云科技众多数据产品的升级中也可以看到。

Long-Term and Forward-Looking Product Strategies: The openness and integrity among products leave room for future technological changes. Sagemaker AI has integrated new generative AI capabilities into the original AI development platform. Given Sagemaker's past success, many developers still favor this platform. In the new GenAI era, instead of discarding the original products, quickly integrating new technologies into them is a huge test of technical and product capabilities for both existing product design and new technology development and new product launches. This characteristic can also be seen in the upgrades of many AWS data products.

数据产品的升级

Upgrades of Data Products

数据作为AI与企业业务优化的关键资产，其重要性仍在不断攀升。在re:Invent 2024大会上，亚马逊云科技再次强调了对数据基础设施的关注，在为企业实施数据战略提供强大的底层支持的同时，也通过不断升级产品功能，进一步优化用户的数据体验。

Data, as the key asset for AI and enterprise business optimization, is growing in importance. At the re:Invent 2024 conference, AWS once again emphasized its focus on data infrastructure. While providing strong underlying support for enterprises to implement data strategies, it also further optimizes the user data experience through continuous product feature upgrades.

优化产品联动能力，持续提升企业实施数据战略过程中的体验：亚马逊云科技提供完善的数据处理与分析工具，为企业在不同阶段的业务需求提供支持。随着生成式 AI 兴起，企业面临的数据流日益复杂，需要多款分析工具与机器学习平台协同配合，才能充分释放数据价值。同时，业务环境的快速变化也对数据处理的敏捷性提出了更高的要求。

Optimizing Product Linkage to Continuously Improve the Enterprise Data Strategy Implementation Experience: AWS offers comprehensive data processing and analysis tools to support enterprises' business needs at different stages. With the rise of generative AI, the data streams that enterprises face are becoming more complex, requiring multiple analysis tools and machine learning platforms to work together to fully unleash data value. At the same time, the rapid changes in the business environment also demand higher agility in data processing.

在新一代的Amazon SageMaker中，SageMaker Unified Studio 通过统一的控制台，显著减少跨平台、跨产品的操作负担，使团队间的协作更加顺畅、模型开发与部署的效率更高。因此，当企业面对来自外部数据与AI训练所产生的大量数据时，能够在确保数据安全与合规的前提下，更高效地处理与分析海量数据，实现从数据到商业洞察再到业务创新的飞轮式增长。

In the new-generation Amazon SageMaker, SageMaker Unified Studio significantly reduces the cross-platform and cross-product operation burden through a unified console, enabling smoother team collaboration and more efficient model development and deployment. Thus, when enterprises face large amounts of data from external sources and AI training, they can process and analyze vast amounts of data more efficiently under the premise of ensuring data security and compliance, achieving a flywheel-like growth from data to business insights and then to business innovation.

此外，亚马逊云科技在 re:Invent 2022 中提出的 zero-ETL理念也正持续演进。在原来Amazon Aurora、Amazon Redshift等亚马逊云科技内部的数据产品实现zero-ETL基础上，如今进一步扩展至对第三方SaaS应用程序的支持。基于此，用户无需构建复杂的数据管道，即可将外部应用的数据直接引入数据产品工具中进行分析，为企业提供更敏捷的数据流转与洞察获取能力。

The zero-ETL concept proposed by AWS at re:Invent 2022 is also evolving. Based on the zero-ETL implementation within AWS's internal data products such as Amazon Aurora and Amazon Redshift, it now further extends to support third-party SaaS applications. As a result, users can directly introduce external application data into data product tools for analysis without building complex data pipelines, providing enterprises with more agile data flow and insight acquisition capabilities.

加强数据治理便捷度，以支持企业识别与维护数据资产：亚马逊云科技推出Amazon SageMaker Lakehouse实现数据湖、数据仓库、运营数据库和企业应用程序的统一管理，提供一致的细粒度访问控制，确保数据治理。同时，通过利用Amazon SageMaker Catalog，用户可以使用生成式AI创建的元数据进行语义搜索，安全地发现、查找和访问数据和模型。另外，Amazon S3新增的 Metadata元数据功能可实现对象元数据自动获取和实时更新，帮助企业更快速地查找业务洞察、实时推理应用程序等所需的数据。

Enhancing Data Governance Convenience to Support Enterprises in Identifying and Maintaining Data Assets: AWS launched Amazon SageMaker Lakehouse to achieve unified management of data lakes, data warehouses, operational databases, and enterprise applications, providing consistent fine-grained access control to ensure data governance. Meanwhile, through Amazon SageMaker Catalog, users can use metadata created by generative AI for semantic searches, safely discovering, finding, and accessing data and models. Additionally, the newly added Metadata feature in Amazon S3 enables automatic acquisition and real-time update of object metadata, helping enterprises find the data required for business insights, real-time inference applications, etc. more quickly.

这些新功能显著提升了数据治理的便捷度，增强数据资产的可见性与可控性，为用户更好地开展数据治理的工作、支持业务决策和创新提供了强大的支撑。

These new features significantly improve the convenience of data governance, enhance the visibility and controllability of data assets, and strongly support users in data governance, business decision-making, and innovation.

数据基础设施持续创新，支撑数据战略可持续实施：在数据库领域，亚马逊云科技推出re:Invent 2024 推出的Amazon Aurora DSQL结合了Amazon TimeSync服务与Serverless的技术能力，提供近乎无限的扩展性和99.999%的多区域可用性。同时，Amazon DynamoDB global tables也新增了多区域强一致性支持，进一步强化其分布式数据库服务能力。

Continuous Innovation in Data Infrastructure to Sustain Data Strategy Implementation: Database Field: Amazon Aurora DSQL, launched at re:Invent 2024, combines the capabilities of Amazon TimeSync service and Serverless technology, providing nearly unlimited scalability and 99.999% multi-region availability. At the same time, Amazon DynamoDB global tables have added multi-region strong consistency support, further strengthening its distributed database service capabilities.

在数据存储领域，亚马逊云科技同样进行了多项细致的优化，例如：Amazon S3增加了HTTP 403访问遭拒信息的额外上下文，以及支持存储浏览器，允许终端用户直接从用户的应用程序中访问Amazon S3的数据；Amazon EBS 支持创建基于时间的快照副本等。

Data Storage Field: AWS has also made multiple detailed optimizations. For example, Amazon S3 has added extra context for HTTP 403 access-denied messages and supports a storage browser, allowing end-users to directly access Amazon S3 data from their applications; Amazon EBS supports creating time-based snapshot copies.

面对企业数据战略的深入推进与 AI 时代的高速发展，亚马逊云科技正持续通过对数据基础设施技术的创新与优化，帮助用户高效识别、管理并释放数据的价值。

Facing the in-depth advancement of enterprise data strategies and the rapid development of the AI era, AWS is continuously innovating and optimizing data infrastructure technologies to help users efficiently identify, manage, and unleash the value of data.

AI产品的升级

Upgrades of AI Products

亚马逊云科技在re:invent 2024大会上推出实用型AI理念，旨在切实满足用户多元化的场景需求。基于此理念亚马逊云科技推出了涵盖基础设施、模型以及应用的全栈联动创新与升级。

At the re:Invent 2024 conference, AWS introduced the practical AI concept, aiming to truly meet the diverse scenario needs of users. Based on this concept, AWS has launched full-stack coordinated innovation and upgrades covering infrastructure, models, and applications.

基础设施层，持续夯实AI计算基座能力：为满足用户日新月异的需求，亚马逊云科技在与英伟达合作推出13个计算实例的基础上，进一步发布了基于NVIDIA Blackwell GPU芯片的P6计算实例，助力亚马逊云科技夯实成为用户使用GPU云服务的绝佳选择。此外，目前的计算实例已可全面采用去年发布的Amazon Trainium2自研芯片，并且经由NeuronLink实现互联的4个Trn2实例、64个Trainium2芯片、最高可达到83.2千万亿次浮点运算的Trn2 UltraServers已预览可用。最后，此次发布全新发布了Trainium3芯片，为下一代生成式AI工作负载打造的3nm工艺、40%能效提升、1倍的性能提升。

Infrastructure Layer, Continuously Strengthening the AI Computing Foundation: To meet users' ever-changing needs, on the basis of launching 13 computing instances in cooperation with NVIDIA, AWS further released the P6 computing instance based on the NVIDIA Blackwell GPU chip, helping AWS solidify its position as an excellent choice for users to use GPU cloud services. In addition, current computing instances can fully adopt the Amazon Trainium2 self-developed chip released last year. The Trn2 UltraServers, with 4 Trn2 instances interconnected via NeuronLink, 64 Trainium2 chips, and a peak performance of 83.2 petaflops, are already available for preview. Finally, the newly released Trainium3 chip, with a 3nm process, 40% energy efficiency improvement, and a 100% performance boost, is designed for next-generation generative AI workloads.

模型应用与开发层，为用户提供模型选择与使用的高自由度：基于“没有一个模型可以适用于所有业务场景”的理念，本次发布会上Amazon Bedrock除了进一步集成业界诸多优质大模型的同时，在功能上也进一步实现了创新升级。首先Amazon Bedrock Model Distillation模型蒸馏功能发布，该功能旨在实现小模型又快又好地在特定场景的使用，相比原始大模型，针对特定用例创建的蒸馏模型不仅速度提升5倍，成本降低75%，而且在RAG等用例中，准确性损失不到2%。此外，防止因模型幻觉而导致的事实性错误，亚马逊云科技推出了发布Amazon Bedrock Automated Reasoning checks自动化推理检查功能，通过交叉引用客户提供的信息以验证模型的响应准确性。最后，亚马逊云科技发布了Amazon Bedrock multi-agent collaboration多智能体协作功能，使得客户能够构建出更加复杂和高效的生成式AI应用。这不仅提升了系统的处理能力，而且也为客户提供了更加灵活和多样化的应用选择。除了Amazon Bedrock的升级，亚马逊云科技重磅推出了新一代基础模型Amazon Nova系列。包括超快速文本生成模型Amazon Nova Micro，能够处理文本、图像和视频并生成文本的多模态模型Amazon Nova Lite、Amazon Nova Pro、Amazon Nova Premier，以及用于生成高质量图像的Amazon Nova Canvas和用于生成高质量视频的Amazon Nova Reel。不仅如此，亚马逊云科技预计将在2025年推出另外两款Amazon Nova模型——“Speech to Speech”和“Any to Any”多模态模型。

Model Application and Development Layer, Offering High Freedom in Model Selection and Use: Based on the idea that "no single model fits all business scenarios", at this conference, Amazon Bedrock not only further integrated many high-quality large models in the industry but also achieved innovative upgrades in functionality.First, the Amazon Bedrock Model Distillation function was released. This function aims to enable small models to be used quickly and well in specific scenarios. Compared to the original large models, the distilled models created for specific use cases not only have a 5x speed increase and a 75% cost reduction but also less than 2% accuracy loss in use cases like RAG.Second, to prevent factual errors caused by model hallucinations, AWS launched the Amazon Bedrock Automated Reasoning checks function, which verifies the accuracy of model responses by cross-referencing information provided by customers.Finally, AWS released the Amazon Bedrock multi-agent collaboration function, enabling customers to build more complex and efficient generative AI applications. This not only improves the system's processing capabilities but also provides customers with more flexible and diverse application options. In addition to the upgrade of Amazon Bedrock, AWS has also launched the new-generation basic model Amazon Nova series, including the ultra-fast text generation model Amazon Nova Micro, the multi-modal models Amazon Nova Lite, Amazon Nova Pro, Amazon Nova Premier that can process text, images, and video and generate text, as well as Amazon Nova Canvas for generating high-quality images and Amazon Nova Reel for generating high-quality video. Moreover, AWS plans to launch two more Amazon Nova models, "Speech to Speech" and "Any to Any" multi-modal models, in 2025.

应用层，助力客户提升作业效率：现今，企业客户普遍面临难以在海量信息中快速找到所需资源的挑战，为改善客户传统的构建体验，简化复杂的数据任务，亚马逊云科技发布了Amazon Q 系列产品推出了创新性功能。首先，Amazon Q Developer于此次会议上展现了性能的大幅提升，在专门用于测试编程能力的SWE bench基准测试中名列前茅，可以解决54.8%的软件开发问题。此外，针对开发人员不同的场景需求，Amazon Q Developer推出了三款自动化智能体，可分别用于实现自动执行单元测试、文档生成和代码审查流程，以进一步提升开发人员效率。

Application Layer, Helping Customers Improve Work Efficiency: Currently, enterprise customers generally face the challenge of quickly finding the required resources in a vast amount of information. To improve customers' traditional construction experience and simplify complex data tasks, AWS has launched innovative features for the Amazon Q series products.First, Amazon Q Developer has shown a significant performance improvement at this conference, ranking high in the SWE bench benchmark test specifically for testing programming capabilities and being able to solve 54.8% of software development problems.Additionally, for different scenario needs of developers, Amazon Q Developer has launched three automated agents, which can be used to implement automatic unit test execution, document generation, and code review processes respectively, further improving developer efficiency.

此次re:invent大会上AI产品的全栈联动创新升级的功能与速度，再次展现出亚马逊云科技的创新实力与对客户尊重负责的态度，印证了活动主题 -“We invent so you can reinvent”。

The full-stack coordinated innovation and upgrade of AI products at this re:Invent conference, in terms of functions and speed, once again demonstrate AWS's innovation strength and its responsible attitude towards customers, confirming the event theme - "We invent so you can reinvent".

对中国用户的价值

Value for Chinese Users

在re:Invent 2024的背景下，亚马逊云科技为中国用户提供了深远的价值，尤其是在全球化运营和本土化支持的双重策略下。

In the context of re:Invent 2024, AWS offers profound value to Chinese users, especially under the dual strategies of global operation and local support.

全球化运营｜亚马逊云科技为中国企业出海和国际企业本地化提供了强有力的支持。

Global Operation｜AWS provides strong support for Chinese enterprises going global and international enterprises localizing in China.

助力中国企业出海：

Helping Chinese Enterprises Go Global:

多区域合规支持：亚马逊云科技通过Amazon Bedrock等服务集成了多种全球领先的大模型，帮助中国企业满足不同区域的数据隐私和合规要求。这对希望拓展欧洲、北美等市场的企业尤为关键。

Multi-Regional Compliance Support: Through services like Amazon Bedrock, AWS integrates multiple world-leading large models, helping Chinese enterprises meet data privacy and compliance requirements in different regions. This is particularly crucial for enterprises looking to expand into markets such as Europe and North America.

生成式AI赋能国际化业务：Amazon Nova系列模型以其高性价比和多模态能力，能够支持文本、图像、视频等多种生成任务，为出海企业在内容生成、客户互动等方面提供强大支持。

Generative AI Empowering International Business: The Amazon Nova series models, with their high cost-performance ratio and multi-modal capabilities, can support multiple generation tasks such as text, image, and video, providing strong support for outbound enterprises in content generation, customer interaction, etc.

统一数据管理与协作：新一代Amazon SageMaker提供统一的数据和AI开发环境，帮助跨国团队协作，提升数据洞察能力，适配不同市场需求。

Unified Data Management and Collaboration: The new-generation Amazon SageMaker provides a unified data and AI development environment, facilitating multinational team collaboration, enhancing data insight capabilities, and adapting to different market needs.

支持外资企业本地化：

Supporting Foreign Enterprises' Localization:

亚马逊云科技在华提供的全栈技术支持（从芯片到应用）使外资企业能够更快适应中国市场需求。例如，通过Amazon Q Developer的新功能，外企可以更高效地完成代码审查、文档生成等任务，从而优化本地研发流程。

The full-stack technical support (from chips to applications) provided by AWS in China enables foreign enterprises to adapt to Chinese market needs more quickly. For example, with the new features of Amazon Q Developer, foreign enterprises can complete tasks such as code review and document generation more efficiently, thus optimizing their local R & D processes.

亚马逊云科技深耕中国市场，与国内开发者和企业合作，通过全国巡展活动推广技术创新，加强与本地生态系统的互动。

AWS has been deeply cultivating the Chinese market, cooperating with domestic developers and enterprises, promoting technological innovation through national tour activities, and strengthening interaction with the local ecosystem.

本土化支持｜亚马逊云科技基于产品和服务打造本地化解决方案，助力行业创新。

Local Support｜AWS builds local solutions based on its products and services to drive industry innovation.

优化成本与性能：

Optimizing Cost and Performance:

Amazon Nova系列模型显著降低了训练和推理成本（最高可节省75%），并提升了推理速度。这种高性价比的解决方案非常适合预算敏感的中小型企业以及初创公司。

The Amazon Nova series models significantly reduce training and inference costs (up to 75% savings) and increase inference speed. This cost-effective solution is very suitable for small and medium-sized enterprises and startups with tight budgets.

基于自研芯片Trainium3的新型计算实例进一步降低了AI工作负载的总拥有成本，为本地企业提供更具竞争力的算力选择。

The new computing instances based on the self-developed Trainium3 chip further reduce the total cost of ownership for AI workloads, providing local enterprises with more competitive computing power options.

贴合本地场景需求：

Meeting Local Scene Requirements:

Amazon Bedrock推出的多智能体协作功能和自动推理检查功能，帮助中国企业在复杂场景中构建更加可靠的生成式AI应用。这些功能特别适用于零售、电商、金融等行业中需要高度个性化和准确性的业务场景。

The multi-agent collaboration and automatic reasoning check functions launched by Amazon Bedrock help Chinese enterprises build more reliable generative AI applications in complex scenarios. These functions are especially applicable to business scenarios in industries such as retail, e-commerce, and finance that require high personalization and accuracy.

针对传统行业现代化转型需求，Amazon Q Developer引入了专门面向大型机和VMware工作负载迁移的工具，加速本地企业数字化转型进程。

For the modernization transformation needs of traditional industries, Amazon Q Developer has introduced tools specifically for migrating mainframe and VMware workloads, accelerating the digital transformation process of local enterprises.

不仅满足中国市场对数据隐私、安全合规的严格要求，还通过优化性能、降低成本和提升效率，为企业在数字化转型过程中提供了强大的技术支持。深度定制的本地服务模式，正在推动包括生命科学、汽车、零售等多个行业的创新与发展。

AWS not only meets the strict requirements of data privacy, security, and compliance in the Chinese market but also provides strong technical support for enterprises in the digital transformation process by optimizing performance, reducing costs, and improving efficiency. The deeply customized local service model is driving innovation and development in multiple industries, including life sciences, automobiles, and retail.

亚马逊云科技通过其全球化资源整合能力和针对中国市场的定制化优化，为中国用户提供了显著价值。从助力出海到赋能本地创新，亚马逊云科技不仅为企业提供了领先技术，还通过降低成本、提升效率以及加强生态合作，为用户创造了实际业务价值。在未来，随着生成式AI技术进一步成熟，这些创新将持续推动中国企业在全球舞台上的竞争力，同时加速各行业数字化转型进程。

Through its global resource integration capabilities and customized optimizations for the Chinese market, AWS provides significant value to Chinese users. From facilitating going global to empowering local innovation, AWS not only offers leading technologies to enterprises but also creates practical business value for users by reducing costs, improving efficiency, and strengthening ecological cooperation. In the future, as generative AI technology further matures, these innovations will continue to boost the competitiveness of Chinese enterprises on the global stage and accelerate the digital transformation process of various industries.

沙利文对未来数据及AI发展的趋势洞察

Frost & Sullivan's Insights into the Future Trends of Data and AI Development

数据与AI结合的未来发展趋势体现在以下三个方面：

The future development trends of the combination of data and AI are reflected in the following three aspects:

数据治理正在走上集成化和智能化的道路：在数据与AI结合的大趋势下，云原生数据治理平台的出现满足了通过AI治理数据的需求。在数据管理方面，AI更趋向于一个高效的附加工具，通过AI的加持可以完成不同数据系统的统一数据管理，从而进一步打破不同数据系统所造成的数据壁垒。这对于混合使用不同数据系统进行存储的企业来说，不但可以进一步降低企业在这方面的管理成本，而且可以方便管理人员进行数据操作。这也进一步达到了AI让数据管理变得简单、高效的目的。

Data Governance Is Moving Towards Integration and Intelligence: Under the big trend of combining data and AI, the emergence of cloud-native data governance platforms meets the need to govern data with AI. In data management, AI is more like an efficient additional tool. With the help of AI, unified data management of different data systems can be achieved, further breaking down the data silos caused by different data systems. For enterprises that use different data systems for storage, this can not only further reduce management costs in this regard but also make it more convenient for managers to operate data. This also further achieves the goal of making data management simple and efficient with AI.

AI正在让数据分析的门槛降低：数据与AI结合的另一个方向就是AI与数据开发工具相结合，在自然语言的输入下，通过AI生成用户所需要的数据分析结果。虽然技术并未做到完全成熟，但是不少厂商已经开始做出该类型产品的尝试。其核心目的是通过AI简化数据处理与AI技术堆栈，从而达到通过AI降低数据分析针对初学者或者零基础人员的使用负担，这种趋势体现出AI正在尝试让数据分析变得易上手、好操作。

AI Is Lowering the Threshold for Data Analysis: Another direction of combining data and AI is the combination of AI and data development tools. With natural language input, AI can generate the data analysis results that users need. Although the technology is not fully mature, many manufacturers have already started to experiment with this type of product. Its core purpose is to simplify data processing and the AI technology stack with AI, so as to reduce the usage burden of data analysis for beginners or those with no foundation. This trend shows that AI is trying to make data analysis more accessible and user-friendly.

数据中心的重要性由于数据与AI的结合变得重要起来：在硬件方面，由于数据与AI的结合，对数据中心的要求也变得严苛起来。数据中心本身需要满足符合AI结合的算力规模与架构、数据管理与疏通、资源池的调度等条件，造成结构较落后且不满足这些条件的数据中心不得不进行一个较大的结构上的改变。数据中心的重要性在这一发展趋势下显现出来，在数据与AI的发展趋势下，如何将数据中心设计的合理高效同时可以存储高增长的数据需求将成为厂商未来需要思考的问题。

The Importance of Data Centers Is Growing Due to the Combination of Data and AI: In terms of hardware, due to the combination of data and AI, the requirements for data centers have become more stringent. Data centers themselves need to meet conditions such as computing power scale and architecture suitable for AI integration, data management and dredging, and resource pool scheduling. Data centers with backward structures that do not meet these conditions have to undergo significant structural changes. The importance of data centers has emerged under this development trend. In the development trend of data and AI, how to design data centers to be reasonable, efficient, and capable of storing data with high growth requirements will be a problem that manufacturers need to consider in the future.

 

 

Frost & Sullivan Independent Research Group