Updated
Updated · Embedded Computing Design · Jun 15
Vecow Unveils 128GB AI Robotics Systems at Automate 2026, Promising 85% Faster Robot Tuning
Updated
Updated · Embedded Computing Design · Jun 15

Vecow Unveils 128GB AI Robotics Systems at Automate 2026, Promising 85% Faster Robot Tuning

1 articles · Updated · Embedded Computing Design · Jun 15

Summary

  • Vecow used Automate 2026 to launch its EAC-7000 robotics supercomputer and EDR-1000 AMR development kit, expanding its industrial AI lineup around NVIDIA-based edge computing.
  • The EAC-7000 runs on NVIDIA Jetson Thor with Blackwell architecture, packing 128GB memory, up to 2,070 FP4 TFLOPS, and 16-channel GMSL sensor inputs for real-time robotic vision and edge inference.
  • The EDR-1000 is built on NVIDIA Nova Orin reference architecture and Vecow’s NAC-1000 core system, adding Bosch 6-axis IMU, IP66 protection, and 5G/4G/WiFi plus M12 PoE+ GigE connectivity for robot-to-cloud links.
  • With Holon Robotics, Vecow also showed an AI-accelerated metal-processing platform using the EVS-3000 and NVIDIA RTX 3500 Ada GPU that cuts robot tuning time by 85% and enables grinding and welding deployment in 15 minutes.
  • The showcase broadens Vecow’s push across industrial automation, with additional Intel- and Qualcomm-based platforms aimed at mobile AI efficiency, intensive computing, and lightweight connected robots.

Insights

NVIDIA's tech powers Vecow's new robots, but can smaller innovators compete in the dawning Physical AI arms race?
With supercomputers now powering humanoid robots, what are the hidden safety and cybersecurity risks for smart factories?
As Physical AI allows 15-minute robot setups, which factory jobs are now most at risk of rapid automation?