Aetina Introduces AIB-AT Series, an Edge AI Platform Based on NVIDIA Jetson Thor for Physical AI and Robotics | RoboticsTomorrow

Aetina Corporation, a main provider of Edge AI strategies, nowadays provided the AIB-AT series, a subsequent-expertise edge AI platform powered by NVIDIA Jetson Thor modules. Designed to fulfill the without warning growing search files from for prime-efficiency, low-latency edge intelligence in rising Physical AI functions, the AIB-AT series marks a significant growth of Aetina’s DeviceEdge portfolio and represents a significant step forward in turning in production-ready computing platforms for Physical AI and robotics.

The AIB-AT series involves the AIB-AT78 powered by NVIDIA Jetson T5000 module and the AIB-AT68 powered by NVIDIA Jetson T4000 module. Constructed on the NVIDIA Blackwell GPU structure, every platforms teach extra special AI compute density in a compact 140 mm × 165 mm build component. This manufacture enables deployment in dwelling-constrained systems equivalent to humanoid robots, self sustaining warehouse vehicles, surgical robots, agricultural drones, and utility inspection robots.

Unparalleled AI Efficiency for Actual-Time Reasoning and Sensor Fusion

The AIB-AT series, leveraging the evolved NVIDIA Jetson Thor platforms, unlocks accurate-time reasoning for general robotics and physical AI with the top stage of efficiency.

With as much as 2070 FP4 TFLOPS of AI compute and 128 GB of memory, the platform delivers as much as 7.5X elevated AI efficiency when put next to the previous expertise, supporting complex workloads equivalent to multimodal perception, sensor fusion, and generative AI inference.

This stage of efficiency permits developers and system integrators to assemble refined robotic systems—together with humanoids, evolved medical units, and self sustaining machines—that require deterministic, low-latency processing at the threshold.

Industrial-Grade Connectivity and Sensor Integration

Aetina has engineered the AIB-AT series specifically for the nerve-racking necessities of business robotics and self sustaining systems. A defining differentiator of the AIB-AT series is its native engineered with twin RJ45 10GbE ports, enabling high-throughput and low-latency AI files alternate, and with one dedicated RJ45 1GbE EtherCAT give a enhance to for deterministic motion retain an eye on, Aetina enables robotics developers to consolidate perception, resolution-making, and retain an eye on into one compact, production-ready system.

Past retain an eye on convergence, the AIB-AT series delivers comprehensive sensor and growth give a enhance to tailored for physical AI workloads. Excessive-velocity QSFP28 connectivity, optimized for NVIDIA Holoscan, enables accurate, low-latency sensor streaming, whereas a 120-pin MIPI interface helps as much as 8x GMSL cameras for evolved multi-glance perception. Additional M.2 growth slots provide PCIe Gen5 NVMe storage, wi-fi connectivity, and optional 4G/5G with 1PPS feature for precision navigation and global time synchronization.

Designed for accurate-world industrial deployment, the platform helps a 9-Forty eight V broad enter voltage fluctuate and operates reliably eventually of -25°C to 80°C, making certain constant efficiency in nerve-racking environments equivalent to factories, warehouses, healthcare amenities, and outdoor self sustaining systems.

Enabling the Subsequent Generation of Physical AI

“Generative physical AI is reworking how machines demand, resolve, and act within the categorical world,” acknowledged Richard Hung, Vice President of Product Division at Aetina, “From self sustaining mobile robots navigating dynamic warehouses to robotic palms adapting grasping in accurate time, and surgical and humanoid robots performing accurate, complex initiatives, physical AI enables extra special adaptability. Our AIB-AT series is motive-constructed to teach the compute density, industrial reliability, and versatile connectivity required for these functions.”

Architected for Subsequent-Generation Physical AI Applications

NVIDIA Jetson Thor represents a brand unusual class of robotic computing platforms, motive-constructed for generative reasoning and multimodal sensor processing. The AIB-AT series helps the corpulent NVIDIA AI’s open models, open-offer frameworks and libraries, together with NVIDIA Isaac, NVIDIA Metropolis, and NVIDIA Holoscan, to boot to agentic AI workflows equivalent to NVIDIA Blueprint for video search and summarization (VSS). From Vision Language Action (VLA) models love the open NVIDIA Isaac GR00T N robotic basis models to open, reasoning VLMs love NVIDIA Cosmos Motive, the AIB-AT series affords developers with a sturdy platform basis, enabling a seamless cloud-to-edge deployment expertise and accelerating the implementation of physical AI innovations.

The AIB-AT series is scheduled to be available by the discontinue of Q1 2026. For more files, please search the recommendation of with www.aetina.com

About Aetina

Aetina, a provider of cutting-edge work AI strategies, affords an excellent quantity of AI computing systems, platforms, hardware, and software tools ultimate for the advent of assorted forms of vertical AI. Aetina’s integrated strategies empower its Arm, x86 computers, GPUs, ASIC hardware, cloud administration software, and development tools with synthetic intelligence, turning in comprehensive, tailored hardware and software suites to allow gentle, hasty system intellectualization growth at the threshold. With its ecosystem community, Aetina makes the advent and deployment of any explicit AI-powered functions extremely achievable by leveraging its global partners’ AI applied sciences. Be taught more at: www.aetina.com