跳至主要內容
機械人

物理AI運算

NVIDIA推出智能機械人專用小型電腦

物理 AI 運算,正透過提供緊湊、高性能的邊緣運算裝置,可直接於裝置本機運行先進 AI 模型,加快智能機械人的部署進程。NVIDIA 全新的 T3000 及 T2000 模組,設計旨在支援人形機械人、自主機械及工業自動化,提供強大的裝置本機處理能力、高效能耗表現,以及簡化開發流程的軟件工具。此平台亦包括 AI 驅動的優化功能,可減少記憶體需求並加快部署速度,讓開發者可打造性能更強的系統,而毋須增加硬件成本。

此做法反映出市場對支援實時決策、而毋須依賴持續雲端連接的邊緣運算需求日益增加。對企業而言,緊湊型 AI 運算平台,可降低部署成本、提升反應速度,並令智能自動化技術,於製造業、物流業、零售業及醫療保健業中,變得更具實用價值。隨著物理 AI 採用率持續增長,可擴展的邊緣運算基礎設施,將成為將自主系統帶入日常商業環境的關鍵基礎。

英文原文

Physical AI computing is accelerating the deployment of intelligent robots by delivering compact, high-performance edge computers capable of running advanced AI models directly on devices. NVIDIA's new T3000 and T2000 modules are designed to support humanoid robots, autonomous machines, and industrial automation with powerful on-device processing, efficient power consumption, and software tools that simplify development. The platform also includes AI-powered optimization features that reduce memory requirements and speed deployment, enabling developers to build more capable systems without increasing hardware costs.

This approach reflects the growing demand for edge computing that supports real-time decision-making without relying on constant cloud connectivity. For businesses, compact AI computing platforms can lower deployment costs, improve responsiveness, and make intelligent automation more practical across manufacturing, logistics, retail, and healthcare. As physical AI adoption grows, scalable edge computing infrastructure will become a key foundation for bringing autonomous systems into everyday commercial environments.

前往原文

物理AI運算
來源
Trend Hunter
發布
2026-07-27
品類
Robots
出處
iottechnews, blogs.nvidia

同品類其他訊號

自動駕駛薄餅外送

Coco Robotics 於部分地區提供熱辣 Little Caesars 薄餅外送服務

機械人