--- title: Nvidia adds Jetson Thor T3000 and T2000 modules for... description: Nvidia has announced two new Blackwell-powered Jetson Thor modules: the T3000 and T2000. The T3000 targets humanoid robots and autonomous mobile robots,... canonical: https://daily.dev/posts/nvidia-adds-jetson-thor-t3000-and-t2000-modules-for-robotics-and-edge-ai-excft2xz5 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Nvidia adds Jetson Thor T3000 and T2000 modules for robotics and edge AI | daily.dev og:description: Nvidia has announced two new Blackwell-powered Jetson Thor modules: the T3000 and T2000. The T3000 targets humanoid robots and autonomous mobile robots,... og:url: https://daily.dev/posts/nvidia-adds-jetson-thor-t3000-and-t2000-modules-for-robotics-and-edge-ai-excft2xz5 og:image: https://api.daily.dev/og/posts/excFT2XZ5.png og:image:alt: Nvidia adds Jetson Thor T3000 and T2000 modules for robotics and edge AI og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Nvidia adds Jetson Thor T3000 and T2000 modules for robotics and edge AI **[Collections](https://daily.dev/sources/collections)** · 3 min read · 0 upvotes · 0 comments ## Summary Nvidia has announced two new Blackwell-powered Jetson Thor modules: the T3000 and T2000. The T3000 targets humanoid robots and autonomous mobile robots, delivering 865 FP4 TFLOPS with 32 GB LPDDR5X memory and 25 GbE connectivity. The T2000 serves as a lower-cost entry point at 400 FP4 TFLOPS with 16 GB memory, aimed at visual AI agents and robotic manipulators. Together they extend the Jetson edge AI range up to 2,000 TFLOPS. Both modules support Nvidia's full Jetson software stack including Cosmos 3 Edge, Isaac GR00T, and Nemotron. New Jetson agent skills automate memory optimization, reportedly saving up to 15 GB per device. The Jetson AGX Thor Developer Kit is available now, T3000 emulation mode arrives in July via JetPack 7.2.1, and both modules ship in Q1 2027. ## Content NVIDIA has expanded its Jetson Thor lineup with two new Blackwell-based modules — the T3000 and T2000 — aimed at bringing the Thor architecture to a broader range of robotics and edge AI applications without the cost and size of the flagship T4000/T5000. ## What the new modules offer The T3000 is the more capable of the two. It delivers 865 FP4 TFLOPS through a 1536-core Blackwell GPU, pairs that with 32GB LPDDR5X memory at 273GB/s bandwidth, and includes 25GbE connectivity. Despite those specs, it's roughly half the physical size of the T5000 — and NVIDIA claims it achieves comparable inference performance for multimodal workloads including LLMs and VLMs. The target use cases are humanoid robots and autonomous systems. The T2000 is the entry point to the Thor family: 400 FP4 TFLOPS, a 1024-core GPU, and 16GB LPDDR5X. It's aimed at visual AI agents, autonomous mobile robots, and manipulators that don't need the full T3000 spec. Together, the two modules extend NVIDIA's Jetson edge AI range from 70 TOPS at the low end up to 2,000 TFLOPS at the top. Both modules support NVIDIA's full software stack, including Isaac GR00T and Nemotron. There's no dedicated dev kit planned for either module — developers can use the existing Jetson AGX Thor Developer Kit, with T3000 emulation mode arriving in July via JetPack 7.2.1. Module availability is scheduled for Q1 2027. ## Cosmos 3 Edge Alongside the hardware, NVIDIA released Cosmos 3 Edge, a 4-billion-parameter open world model built for on-device inference on memory-constrained hardware like Jetson Thor, RTX GPUs, and DGX systems. The architecture is a bit unusual: it combines two transformer towers over a shared multimodal representation — an autoregressive tower for vision and text reasoning, and a diffusion tower for video, audio, and action generation. It runs at robot-control resolution (640×360), generates 32 actions per inference, and hits 15 Hz real-time control on Jetson Thor. NVIDIA says it ranks first on VANTAGE-Bench among 4B-parameter models. A policy variant post-trained on the DROID dataset for pick-and-place tasks is also available, along with training scripts. There's also a 4-step distillation checkpoint called Cosmos 3 Super that cuts inference steps from 35–50 down to 4, for up to 25× speedup. vLLM added day-0 support for Cosmos 3 Edge through both vLLM and vLLM Omni, so it can be served immediately. ## Jetson agent skills NVIDIA also released Jetson agent skills — tools that automate memory optimization across the Jetson portfolio. In real-world deployments, NVIDIA reports up to 15GB in memory savings, which matters a lot when you're running large models on edge hardware with fixed memory budgets. ## Similar posts on daily.dev - [Accelerate AI Inference for Edge and Robotics with NVIDIA Jetson T4000 and NVIDIA JetPack 7.1](https://daily.dev/posts/accelerate-ai-inference-for-edge-and-robotics-with-nvidia-jetson-t4000-and-nvidia-jetpack-7-1-lxnil3vmy) · NVIDIA Developer · 1 upvotes · 0 comments - [Aetina Adds Jetson T3000 and T2000 Support to Edge AI Systems](https://daily.dev/posts/aetina-adds-jetson-t3000-and-t2000-support-to-edge-ai-systems-sh4oxkhl7) · Embedded.com · 0 upvotes · 0 comments - [NVIDIA IGX Thor Robotics Processor Brings Real-Time Physical AI to the Industrial and Medical Edge](https://daily.dev/posts/nvidia-igx-thor-robotics-processor-brings-real-time-physical-ai-to-the-industrial-and-medical-edge-j8y6svy6x) · NVIDIA · 0 upvotes · 0 comments --- Tags: [#ai](https://daily.dev/tags/ai), [#embedded](https://daily.dev/tags/embedded), [#local-ai](https://daily.dev/tags/local-ai), [#nvidia](https://daily.dev/tags/nvidia), [#robotics](https://daily.dev/tags/robotics) [View this post on daily.dev](https://daily.dev/posts/nvidia-adds-jetson-thor-t3000-and-t2000-modules-for-robotics-and-edge-ai-excft2xz5) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Nvidia adds Jetson Thor T3000 and T2000 modules for robotics and edge AI","url":"https://daily.dev/posts/nvidia-adds-jetson-thor-t3000-and-t2000-modules-for-robotics-and-edge-ai-excft2xz5","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/nvidia-adds-jetson-thor-t3000-and-t2000-modules-for-robotics-and-edge-ai-excft2xz5"},"datePublished":"2026-07-15T23:34:24.388Z","dateModified":"2026-10-15T02:16:55.190Z","description":"Nvidia has announced two new Blackwell-powered Jetson Thor modules: the T3000 and T2000. The T3000 targets humanoid robots and autonomous mobile robots,...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/6b285e9c2f6169b4049e782140dc814f?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/6b285e9c2f6169b4049e782140dc814f?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Collections","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Collections","logo":"https://media.daily.dev/image/upload/s--fk_6ycEi--/f_auto,q_auto/v1780996001/logos/collections?_a=BAMAMiWQ0","url":"https://daily.dev/sources/collections"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/nvidia-adds-jetson-thor-t3000-and-t2000-modules-for-robotics-and-edge-ai-excft2xz5","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai,embedded,local-ai,nvidia,robotics","timeRequired":"PT3M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Collections","item":"https://daily.dev/sources/collections"},{"@type":"ListItem","position":3,"name":"Nvidia adds Jetson Thor T3000 and T2000 modules for robotics and edge AI"}]} ```