«

英伟达与达索系统将共建工业人工智能平台。

qimuai 发布于 阅读:2 一手编译


英伟达与达索系统将共建工业人工智能平台。

内容来源:https://aibusiness.com/industrial-manufacturing/nvidia-dassault-build-industrial-ai-platform

内容总结:

英伟达与达索系统强强联手 推出工业AI平台推动产业变革

在今日举行的达索系统年度设计与工程盛会“3D体验世界”大会上,芯片巨头英伟达与全球领先的软件公司达索系统共同宣布,将深化合作,携手打造一个全新的工业人工智能平台。该平台旨在融合达索系统的虚拟孪生技术与英伟达的加速AI基础设施,为制造业、工程与生命科学等领域的复杂系统,提供大规模仿真、设计与应用支持。

根据合作规划,达索系统的3D体验平台及其虚拟孪生技术,将与英伟达的AI基础设施、开放模型及加速软件库深度集成,共同构建科学验证的“行业世界模型”。达索系统研发与创新执行副总裁弗洛朗斯·胡在发布会前简报中阐释,虚拟孪生是一种科学的、多学科、多尺度的数字化表征,能在实体诞生前于真实条件下进行全方位测试。她以航空发动机设计为例指出,虚拟孪生不仅能构建三维模型,更能模拟真实飞行条件下的热力学、应力、载荷与性能,从而在物理原型出现之前即可验证产品可行性、预测维护需求并测试设计变更。

该平台的另一核心亮点是引入“虚拟助手”——内嵌于3D体验平台中的智能体AI系统。胡副总裁将其描述为能够“理解意图、基于行业世界模型进行推理并协调行动”的AI专家。目前,平台已开发出面向商业的Aura、面向工程任务的Leo以及面向科学应用的Marie三款助手模型。她强调,这一转变标志着从“软件即服务”迈向“智能体即服务”,而英伟达的AI模型与编排工具是实现这一变革的关键。

“没有大规模的AI工厂,就不会有虚拟孪生工厂、行业世界模型和虚拟助手,”胡副总裁总结道,“此次合作不仅是技术整合,更是关于AI如何通过科学与经过验证的可靠智能来变革工业的共同愿景。”

英伟达创始人兼首席执行官黄仁勋在新闻稿中表示:“我们与达索系统一同,正在改变数百万研究人员、设计师和工程师构建世界最大产业的方式。”此次合作也体现了英伟达向物理AI领域拓展的战略布局。该公司将采用达索系统的基于模型的系统工程方法来设计自身的AI工厂,首先应用于英伟达Rubin平台,并整合至其Omniverse Blueprint中,以实现大规模AI工厂的部署。

此次强强联合,旨在为企业提供高质量的仿真环境,从根本上变革行业构建方式,标志着工业设计与生产智能化进程迈入新阶段。

中文翻译:

由谷歌云赞助
选择您的首个生成式AI应用场景
要开启生成式AI之旅,首先应关注能够提升人类信息交互体验的领域。

该平台将为企业提供高质量仿真方案,从而"彻底重塑"各行业的构建模式。

英伟达与软件公司达索系统正携手打造工业人工智能平台,将虚拟孪生技术与加速AI基础设施深度融合。双方在达索系统年度设计与工程盛会"3D体验世界"大会上宣布了此项合作。

根据合作协议,达索系统的3D体验平台及虚拟孪生技术将与英伟达的AI基础设施、开源模型及加速软件库进行整合,共同创建两家公司所称的经过科学验证的"行业世界模型"。

这些模型旨在支持制造业、工程学和生命科学领域的大规模复杂系统仿真、设计与应用。

"虚拟孪生是一种科学的、多学科、多尺度的数字化呈现,能在实体诞生前完全模拟真实环境下的测试场景,"达索系统研发与创新执行副总裁弗洛朗丝·胡在预发布会上表示。

胡指出,这项技术使企业能够在实体部署前对产品进行全方位验证,实现从项目初始阶段推动产业转型。

"对于设计新型发动机的航空航天制造商而言,虚拟孪生不仅是三维模型。它能模拟真实飞行条件下的热力学、应力、载荷与性能表现,"她补充道,"在实体原型尚未诞生时,即可预判其运行效能。虚拟世界还能预测维护需求并测试各类设计变更。"

该平台另一核心特性是引入"虚拟助手"——嵌入3D体验平台的智能体AI系统。胡表示这些助手如同AI专家,能够"理解用户意图,基于行业世界模型进行推理并协调各项操作"。

目前已有三款模型投入应用:面向商业场景的奥拉、工程任务的利奥以及科学用例的玛丽。

胡将这一变革描述为从软件即服务向"智能体即服务"的演进,其实现依托英伟达的AI模型与协调工具,堪称颠覆性创新。

"若没有规模化AI工厂,就不会有虚拟孪生工厂、行业世界模型和虚拟助手,"她强调,"此次合作不仅是技术整合,更是对AI如何通过科学与可信智能推动产业转型的共同愿景。"

此次深化合作也折射出英伟达向实体AI领域的战略拓展。

该公司正采用达索系统的基于模型的系统工程方法,从英伟达鲁宾平台起步设计自有AI工厂,并计划整合至英伟达Omniverse蓝图以实现大规模AI工厂部署。

"我们正与达索系统共同改变数百万研究人员、设计师和工程师构建全球最大规模产业的方式,"英伟达创始人兼首席执行官黄仁勋在新闻声明中表示。

英文来源:

Sponsored by Google Cloud
Choosing Your First Generative AI Use Cases
To get started with generative AI, first focus on areas that can improve human experiences with information.
The platform will offer companies high-quality simulations to "transform" how industries are built from the ground up.
Nvidia and software company Dassault Systèmes are partnering to build an industrial AI platform, combining virtual twin technology with accelerated AI infrastructure.
The partnership was announced today at 3DExperience World, Dassault Systèmes' annual design and engineering event.
Under the collaboration, Dassault Systèmes' 3DExperience platform and virtual twin technologies will be integrated with Nvidia's AI infrastructure, open models and accelerated software libraries to create what the companies described as science-validated "industry world models."
The models are designed to support large-scale simulation, design and use of complex systems across manufacturing, engineering and life sciences.
"A virtual twin is a scientific, multidisciplinary and multiscale representation that is fully testable under real conditions before anything exists," Dassault Systèmes' executive vice president of research and innovation Florence Hu said during a pre-briefing.
Hu said this capability allows companies to holistically validate products before physical deployment and enable industrial transformation from a project's inception.
"For an aerospace manufacturer designing a new engine, the virtual twin is not just a 3D model. It simulates thermal dynamics, stress, loads and performance under real flight conditions," she added. "Before a single physical prototype exists, you can know whether it will work and the virtual world allows you to predict maintenance and test any modification."
Another central feature of the platform is the introduction of "virtual companions," agentic AI systems embedded in the 3DExperience platform. Hu said these companions act as AI experts that "understand intent, reason with industry world models and orchestrate actions."
Three of these models have already been developed: Aura for business, Leo for engineering tasks and Marie for scientific use cases.
Hu described the shift as a move from SaaS to "agents as a service," enabled by Nvidia's AI models and orchestration tools, and is pitched as a transformative tool.
"Without AI factories at scale, there are no virtual twin factories, no industry world models and no virtual companions," she said. "This partnership is not just a technology integration, it's a shared vision for how AI will transform industry through science and validated, trustworthy intelligence."
The expanded collaboration also reflects Nvidia’s push into physical AI.
The company is adopting Dassault Systèmes' model-based systems engineering approach to design its own AI factories, starting with the Nvidia Rubin platform and integrating into Nvidia's Omniverse Blueprint for large-scale AI factory deployment.
"Together with Dassault Systèmes, we're transforming how millions of researchers, designers and engineers build the world's largest industries," Nvidia founder and CEO Jensen Huang said in a press release.

商业视角看AI

文章目录


    扫描二维码,在手机上阅读