Nvidia 拟以至少 5 亿美元收购扩散语言模型初创公司 Inception
Nvidia bets at least $500 million on Inception's diffusion models to strengthen its open-source Nemotron lineup
Nvidia 要花至少 5 亿美元买下 Inception,就是那家用扩散方式做文本生成的公司,不按 token 一个个蹦,生成更快。它会把这些技术塞进开源的 Nemotron 系列,还用于机器人。做开源模型或关注非自回归路线的开发者应该看看。
据 The Information 报道,Nvidia 正接近以至少 5 亿美元收购成立仅两年的 AI 初创公司 Inception。Inception 的核心是基于扩散过程构建语言模型,与传统 ChatGPT 类模型逐个 token 生成的方式不同,扩散方法可以并行生成文本,在推理速度上具备优势。
这笔收购将直接补强 Nvidia 自家的开源 Nemotron 模型系列,为其开源模型路线引入扩散语言模型的技术能力。报道还指出,Nvidia 计划将扩散模型能力用于其机器人业务方向,与其硬件生态形成协同。
对行业而言,这是 GPU 厂商首次大额押注扩散式文本生成路线。相比 GPT-4 等主流自回归模型,扩散语言模型目前生态尚不成熟,Nvidia 的入局可能加速这条路线的开源进展,值得关注与 AI2、Inception Labs 等扩散模型开源项目的竞争关系。
Nvidia bets at least $500 million on Inception's diffusion models to strengthen its open-source Nemotron lineup
Nvidia is close to acquiring AI startup Inception for at least $500 million, according to The Information. The company is just two years old and builds language models based on diffusion processes, an alternative approach to the token-by-token generation used by models like ChatGPT. The deal would bolster Nvidia's own open-source models and its push into robotics. The article Nvidia bets at least $500 million on Inception's diffusion models to strengthen its open-source Nemotron lineup appeared first on The Decoder .
This report is categorized under AITOP's "Industry" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Nvidia, Inception, Nemotron, diffusion models.
Source attribution: curated from Decoder (@Maximilian Schreiner). Cross-source corroboration: 11 independent sources have covered this development (including NVIDIA AI, marktechpost, Naval), confirming industry-wide signal strength.