内容来源：https://www.wired.com/story/nvidia-becomes-major-model-maker-nemotron-3/

内容总结：

芯片巨头英伟达近日发布了一系列前沿开源模型“Nemotron 3”，并配套开放训练数据与定制工具，标志着其在巩固AI芯片霸主地位的同时，正积极向模型研发领域纵深布局。此举被视为应对OpenAI、谷歌等AI公司自主研发芯片趋势的战略对冲，旨在通过开源生态增强开发者黏性。

当前，开源模型已成为全球AI创新的关键基础设施。尽管美国企业普遍转向技术封闭，中国公司却持续推动开源生态繁荣——根据开源平台Hugging Face数据，百度、阿里、深度求索等中国企业发布的开源模型因更新频繁、技术透明，更受开发者青睐。英伟达此次不仅公开了训练数据，还推出支持强化学习的工具库，其采用的混合专家架构特别适用于开发能执行计算机操作的AI智能体。

值得关注的是，地缘政治因素正为英伟达带来挑战。尽管美国政府近期批准其向中国出口H200芯片，但中国推动技术自主的举措已促使本土企业更多采用国产芯片。若中国AI模型与国产芯片形成深度绑定，或将对英伟达的市场地位产生长远影响。

业内分析指出，开源战略已成为AI竞争的重要维度。英伟达副总裁卡里·安·布里斯基强调，开源能加速AI创新并赋能全球经济，而Meta等企业却逐步收紧开源政策。这种分化格局下，技术透明度与生态开放性或将成为影响全球AI发展走向的关键变量。

中文翻译：

英伟达通过向人工智能公司供应芯片赚得盆满钵满，如今这家芯片制造商朝着成为更专业的模型开发商迈出关键一步——发布了一系列尖端开源模型，并配套提供帮助工程师使用这些模型的数据和工具。

当前正值OpenAI、谷歌和Anthropic等人工智能公司竞相开发自主芯片之际，此举可视为英伟达应对这些企业逐渐脱离其技术生态的未雨绸缪之举。

开源模型已成为人工智能生态系统的关键组成部分，众多研究机构和初创企业正利用它们进行实验、原型开发和产品构建。尽管OpenAI和谷歌提供小型开源模型，但其更新频率远不及中国竞争对手。开源项目托管平台Hugging Face数据显示，基于此等多重因素，中国企业开发的开源模型目前更受市场青睐。

根据英伟达发布前公布的基准测试分数，其全新Nemotron 3系列模型堪称当前可下载、可修改、可在自有硬件运行的最佳模型之一。"开源创新是人工智能进步的基石，"首席执行官黄仁勋在官方声明中表示，"通过Nemotron，我们将尖端人工智能转化为开放平台，为开发者提供构建规模化智能体系统所需的透明度与效率。"

相较于多数美国竞争对手，英伟达采取了更彻底的开放策略——公开Nemotron模型的训练数据集，此举将显著降低工程师修改模型的难度。公司还同步发布了定制化与微调工具，包括新型混合潜在专家模型架构。英伟达表示该架构特别适用于构建能在计算机或网络环境中执行操作的智能体，同时推出支持用户通过模拟奖惩机制（强化学习）训练智能体执行任务的算法库。

Nemotron 3系列提供三种规格：300亿参数的Nano版、1000亿参数的Super版及5000亿参数的Ultra版。模型参数规模大致对应其能力强度与运行复杂度，最大规格模型需依赖昂贵硬件集群才能运转。

模型基石
英伟达企业级生成式AI软件副总裁卡里·安·布里斯指出，开源模型对AI开发者至关重要的三大原因在于：开发者日益需要针对特定任务定制模型；将查询任务分配给不同模型往往能提升效率；通过模拟推理训练可更有效地激发模型产生智能响应。"我们相信开源是AI创新的根基，将持续推动全球经济发展，"布里斯强调。

社交媒体巨头Meta于2023年2月以Llama为名发布了首批先进开源模型。但随着竞争白热化，Meta已暗示未来版本可能不再开源。

这折射出人工智能行业的宏观趋势。过去一年间，美国企业逐渐背离开放原则，对研究成果日趋保密，更不愿向竞争对手透露最新工程技术。

模型集成平台OpenRouter近期报告显示，2025年其系统处理的所有文本数据单元中，开源模型处理量占比约三分之一。深度求索、阿里巴巴、月之暗面、智谱AI、MiniMax等中国企业持续发布强力开源模型，并公布研究进展细节，使其产品对工程师的实验吸引力与日俱增。

这对英伟达可能构成潜在挑战。该公司硬件在AI领域已重要到其芯片成为特朗普对华贸易谈判的筹码。美国政府近期虽允许英伟达向中国出口前代旗舰产品H200芯片，但中国政府着力推动技术自主，正采取措施鼓励企业采用国产芯片。这种趋势可能导致中国AI模型与国产芯片深度绑定，进而动摇英伟达的市场地位。

英文来源：

Nvidia has made a fortune supplying chips to companies working on artificial intelligence, but today the chipmaker took a step toward becoming a more serious model maker itself by releasing a series of cutting-edge open models, along with data and tools to help engineers use them.
The move, which comes at a moment when AI companies like OpenAI, Google, and Anthropic are developing increasingly capable chips of their own, could be a hedge against these firms veering away from Nvidia’s technology over time.
Open models are already a crucial part of the AI ecosystem with many researchers and startups using them to experiment, prototype, and build. While OpenAI and Google offer small open models, they do not update them as frequently as their rivals in China. For this reason and others, open models from Chinese companies are currently much more popular, according to data from Hugging Face, a hosting platform for open source projects.
Nvidia’s new Nemotron 3 models are among the best that can be downloaded, modified, and run on one’s own hardware, according to benchmark scores shared by the company ahead of release.
“Open innovation is the foundation of AI progress,” CEO Jensen Huang said in a statement ahead of the news. “With Nemotron, we’re transforming advanced AI into an open platform that gives developers the transparency and efficiency they need to build agentic systems at scale.”
Nvidia is taking a more fully transparent approach than many of its US rivals by releasing the data used to train Nemotron—a fact that should help engineers modify the models more easily. The company is also releasing tools to help with customization and fine-tuning. This includes a new hybrid latent mixture-of-experts model architecture, which Nvidia says is especially good for building AI agents that can take actions on computers or the web. The company is also launching libraries that allow users to train agents to do things using reinforcement learning, which involves giving models simulated rewards and punishments.
Nemotron 3 models come in three sizes: Nano, which has 30 billion parameters; Super, which has 100 billion; and Ultra, which has 500 billion. A model’s parameters loosely correspond to how capable it is as well as how unwieldy it is to run. The largest models are so cumbersome that they need to run on racks of expensive hardware.
Model Foundations
Kari Ann Briski, vice president of generative AI software for enterprise at Nvidia, said open models are important to AI builders for three reasons: Builders increasingly need to customize models for particular tasks; it often helps to hand queries off to different models; and it is easier to squeeze more intelligent responses from these models after training by having them perform a kind of simulated reasoning. “We believe open source is the foundation for AI innovation, continuing to accelerate the global economy,” Briski said.
The social media giant Meta released the first advanced open models under the name Llama in February 2023. As competition has intensified, however, Meta has signaled that its future releases might not be open source.
The move is part of a larger trend in the AI industry. Over the past year, US firms have moved away from openness, becoming more secretive about their research and more reluctant to tip off their rivals about their latest engineering tricks.
A recent report from OpenRouter, a company that gives people access to different models through a single user interface, shows that open models accounted for around a third of all tokens—units of text and other data—sent through its systems in 2025. Chinese firms including DeepSeek, Alibaba, Moonshot AI, Z.ai, and MiniMax regularly release powerful open models and publish details about their research advancements that make their offerings more appealing for engineers to experiment with.
This could prove troublesome for Nvidia. The company’s hardware has become so important in the world of AI that its silicon has become a bargaining chip in Trump’s trade dealings in China. The US government recently said it would allow Nvidia to export H200 chips—the best of its previous generation—to China, but the Chinese government is keen to achieve greater technological independence and has taken steps to push Chinese companies to use home-grown chips. This could mean Chinese AI models become more closely aligned with Chinese silicon, which could undermine Nvidia’s position.