Aetina Corporation has announced the availability of its new DeviceEdge AIE-KT78 and AIE-KT68 edge AI systems. These systems are built around NVIDIA's Jetson Thor technology, known for its advanced AI capabilities.
The integration of NVIDIA Blackwell architecture brings enhanced AI compute power, high-bandwidth sensor connectivity, and deterministic industrial control. These features make the systems suitable for Collaborative Robots (Cobots), humanoid robots, and emerging autonomous machinery, offering developers the capability to process sensory data, reason, plan, and execute actions effectively in real-world settings.
Performance and capability
As robotics transition towards context-aware, autonomous functions, it becomes crucial for systems to handle numerous high-resolution sensor inputs, operate extensive multimodal models, and maintain real-time control. The AIE-KT78, featuring the NVIDIA Jetson T5000 module with 128GB of 256-bit LPDDR5X memory, efficiently delivers up to 2,070 FP4 TFLOPS of AI performance. Meanwhile, the AIE-KT68, powered by the NVIDIA Jetson T4000 module with 64GB of similar memory, offers up to 1,200 FP4 TFLOPS—balancing power, performance, and cost efficiency in tackling rising compute demands.
Both the AIE-KT78 and AIE-KT68 systems allow for local execution of complex AI models such as multimodal generative AI and Vision-Language-Action (VLA) models, which minimises cloud dependency and enhances data control. These models enable robots to interpret surroundings and react instantly. Equipped with QSFP28 connectivity providing up to 4x25Gbps bandwidth and dual RJ45 10GbE ports, the systems support the processing of massive data streams from various high-resolution sensors, essential for a comprehensive understanding and perception of environments.
Autonomous system architecture
Featuring a dedicated RJ45 1GbE EtherCAT port, these systems serve as an EtherCAT Master
Featuring a dedicated RJ45 1GbE EtherCAT port, these systems serve as an EtherCAT Master, ensuring precise AI inference synchronisation with motors, sensors, and actuators. This integration augments the adaptability and safety of cobots, while enhancing the intricate spatial awareness and ongoing motion decisions necessary for humanoid robots. Additionally, it simplifies architecture by reducing the need for extra hardware, beneficial for advanced autonomous machines, cobots, and industrial robot arms advancing from concept stages to practical deployment.
The innovative design of the AIE-KT78/68 systems incorporates high compute performance, industrial-grade interfaces, and efficient thermal management within an 80mm compact frame. They operate within a wide voltage range from 9 to 48VDC, and can withstand temperatures ranging from -25°C to +55°C. The inclusion of USB 3.2, isolated digital I/O, and M.2 expansion slots supports versatile configurations. Running on Linux, the systems leverage the extensive NVIDIA AI software suite, aiding developers in optimising and deploying various AI and robotics applications.
Industry impact
Richard Hung, Vice President of Product Division at Aetina, noted the shift in robotics competitive dynamics: "The competitive edge for cobots and humanoid robots has shifted from the compute performance of a single model to whether a system can integrate perception, reasoning, decision-making, and action in real time in real-world environments."
As an NVIDIA Elite Partner, Aetina consolidates its edge AI engineering expertise, providing comprehensive technical support, which enables smoother transition from initial conceptualisation to scaled deployment, enhancing developers’ and integrators’ capacity to maximise their automation investments.
Aetina Corporation, a pioneer AI solution provider and an accelerator of AI infrastructure, announces the general availability of its DeviceEdge AIE-KT78 and AIE-KT68 flagship edge AI systems.
Powered by NVIDIA Jetson Thor, both new systems integrate NVIDIA Blackwell architecture's high-performance AI compute, high-bandwidth sensor connectivity, and deterministic industrial control into a compact edge system — purpose-built for Collaborative Robots (Cobots), humanoid robots, and next generation autonomous machines — helping developers perceive, reason, decide, and act in unison in real world environments.
High-resolution sensor streams
As robots evolve from fixed-routine automation toward context-aware, autonomous decision-making, systems must simultaneously process multiple high-resolution sensor streams, run large multimodal models, and maintain real-time, deterministic control over motors, joints, and external actuators.
Facing this rapidly escalating compute demand, existing edge platforms delivering only a few hundred TOPS are struggling to keep up. AIE-KT78 is powered by the NVIDIA Jetson T5000 module with 128GB of 256-bit LPDDR5X memory, delivering up to 2,070 FP4 TFLOPS of AI performance; AIE-KT68 is powered by the NVIDIA Jetson T4000 module with 64GB of 256-bit LPDDR5X memory, delivering up to 1,200 FP4 TFLOPS — striking a balance between performance, power consumption, and system cost.
Strengthening data control
Both systems support on-device execution of multimodal generative AI, large language models, vision-language models, and Vision-Language-Action (VLA) models, reducing cloud latency and strengthening data control.
The AIE-KT78/68 enables robots to run advanced VLA models locally, integrating visual perception, language understanding, and action generation into a single decision-making pipeline that interprets environmental context and responds in real time. To handle such data-intensive perception workloads, the AIE-KT Series simultaneously integrates QSFP28 connectivity (delivering up to 4x25Gbps of bandwidth depending on configuration) and dual RJ45 10GbE ports, enabling high-speed processing of massive real time data streams from high-resolution cameras, 3D LiDAR, radar, depth cameras, IMUs, and various industrial sensors — while supporting up to 8 GMSL2 camera inputs to provide the high-bandwidth data foundation robots need for multi-view environmental perception and spatial understanding.
Advanced autonomous machines
A dedicated RJ45 1GbE EtherCAT port functions as an independent EtherCAT Master, using microsecond level synchronisation to tightly connect high-level AI inference with underlying motors, joints, sensors, and actuators — consolidating perception, decision-making, and deterministic control within a single system.
For collaborative robots, this architecture enhances the flexibility to safely work alongside people and quickly switch between tasks; for humanoid robots, it supports more complex spatial awareness and continuous motion decision-making. It also reduces the need for additional control hardware and cross-platform integration, simplifying the system architecture for cobots, humanoid robots, industrial robotic arms, and advanced autonomous machines — helping businesses accelerate the transition from proof-of-concept to real-world deployment.
Deploying multimodal inference
The AIE-KT78/68 integrates compute performance, industrial interfaces, and thermal design into a compact system just 80mm thick, supporting a wide 9–48VDC input range and stable operation across −25°C to +55°C.
The system is equipped with USB 3.2, isolated digital I/O, and M.2 expansion slots. It runs Linux with the full NVIDIA AI software stack, including NVIDIA JetPack 7, NVIDIA CUDA, NVIDIA TensorRT, NVIDIA DeepStream, NVIDIA Holoscan, and NVIDIA Isaac ROS, helping developers optimise and deploy multimodal inference, computer vision, sensor fusion, and robotics applications.
Fully validated peripheral ecosystem
“The competitive edge for cobots and humanoid robots has shifted from the compute performance of a single model to whether a system can integrate perception, reasoning, decision-making, and action in real time in real-world environments,” said Richard Hung, Vice President of Product Division at Aetina.
“The AIE KT78/68 transforms the powerful compute capability of NVIDIA Jetson Thor into a compact, mass production-ready system. As an NVIDIA Elite Partner, we combine over a decade of edge AI engineering experience with a fully validated peripheral ecosystem, providing end-to-end technical support — from mechanical design and thermal engineering to BSP and system integration — to help developers and system integrators reduce integration risk and total cost of ownership, accelerate the path from proof-of-concept to scaled deployment, and further realise the return on their automation investment.”
