Road to embedded world '23: New Taipei City, Taiwan, Aetina Corporation

January 20, 2023

Blog

In literature when one crosses a body of water, it tends to symbolize a creation or a rebirth. Today on our Road to embedded world ’23 we are crossing a few bodies of water and making our way to the Xizhi District, New Taipei City, Taiwan. Here we are taking a tour of Aetina and learning more about its latest creations hitting the conference floor.

Aetina will introduce its latest offerings in Edge AI Solutions with demonstations showcasing how it leverages the power of NVIDIA to create the next generation of AI systems and hardware.

Make sure to stop by Aetina’s Booth, Hall 1-370, to view the following solutions for your AI system needs:

AI Accelerator and GPU modules for use in AI and IoT

The modules can accelerate the inference process for high-level AI-powered systems. These modules include: 

• M3A1000-PP: MXM module powered by NVIDIA RTX A1000 Embedded GPU, providing 6.66 TFLOPS of AI performance (product page: https://www.aetina.com/products-detail.php?i=515)
• M3A2000-VY: MXM module powered by NVIDIA RTX A2000 Embedded GPU, providing 8.25 TFLOPS of AI performance (product page: https://www.aetina.com/products-detail.php?i=516)
• M3A4500-WP: MXM module powered by NVIDIA RTX A4500 Embedded GPU, providing 18.55 TFLOPS of AI performance (product page: https://www.aetina.com/products-detail.php?i=517)
• AI-MXM-H84A: MXM module powered by Hailo-8™ AI inference processors, providing up to 104 TOPS of AI performance (product page: https://www.aetina.com/products-detail.php?i=556)
• M3A500-PP: MXM module powered by NVIDIA RTX A500 Embedded GPU, providing 6.54 TFLOPS of AI performance 
• AI-MXM-S41A: MXM module powered by Sophgo BM1684 AI processor, providing 17.6 TOPS of AI performance 

Edge Computing Systems and Platforms for AI Inference:

• ASIC-based edge computing systems: embedded computers that are powered by system-on-modules made with ASIC chips
• AIE-CP1A-A1: edge AI box powered by Blaize® Pathfinder P1600 embedded system-on-module, providing up to 16 TOPS of AI performance (product page: https://www.aetina.com/products-detail.php?i=560)
• NVIDIA Jetson-based edge computing systems and platforms: embedded computers powered by NVIDIA Jetson modules, including the latest AGX Orin, and upcoming Orin NX and Orin Nano
• AIB-MX12: edge computing platform powered by an NVIDIA Jetson AGX Orin system-one-module (carrier board paired with an Jetson AGX Orin SoM; product page: https://www.aetina.com/products-detail.php?i=511)
• AIB-MX13: edge computing platform powered by an NVIDIA Jetson AGX Orin system-one-module (carrier board paired with an Jetson AGX Orin SoM; product page: https://www.aetina.com/products-detail.php?i=512)
• AIE-KX12: edge computing box by an NVIDIA Jetson AGX Orin system-on-module (product page: https://www.aetina.com/products-detail.php?i=513)
• AIE-KX13: edge computing boxes powered by an NVIDIA Jetson AGX Orin (product page: https://www.aetina.com/products-detail.php?i=514)
• AIB-SN31/41: edge computing platform powered by an NVIDIA Jetson Orin NX system-on-module (*product page not live yet)
• Mini M1 (AIE-CN11/21 | AIE-CT41): palm-size, energy-efficient box PC powered by an NVIDIA Jetson Xavier NX or Jetson TX2 NX SoM (product page: https://www.aetina.com/products-detail.php?i=486)
• Systems and platforms powered by NVIDIA Jetson Orin NANO SoM (*product page not live yet)
• Fanless embedded computers (TBD)
• x86 AI inference platforms: Embedded computers that feature an Intel CPU and are compatible with AI accelerator and GPU modules for performance boost
• AIP-SQ67: x86 AI inference platform featuring an Intel® 12th Gen Core™ i7/i5 processor and compatible with an MXM module for AI performance boost (datasheet: https://www.aetina.com/wp-content/uploads/Datasheet/Aetina_AIP-SQ67_datasheet.pdf) 
• AIP-FR68 & AIP-CR68: x86 inference platforms that feature an Intel® 13th Core i5/i7 processor and are compatible with extra GPU or AI accelerator modules for AI performance boost (*product page not live yet)

AI Software Tools:

• Tools for Edge Device Management from Cloud and AI Model PoC & Inference Running
• Aetina Triton Inference Server: a software tool for AI model performance online evaluation and AI inference running in any web browser (*product page not live yet).
• Aetina EdgeEye: a software tool that enable users to monitor the real-time running status of edge devices located in different places and manage them (including remote device reboot) in any web browser on one central computer (*product page not live yet)

Solution Live Demos:

License plate recognition solution in cities: the solution involves an AI model that can analyze vehicle images and videos received by cameras and recognize the plate numbers. The model is deployed/installed on Aetina’s edge computing box—AIE-KX13, powered by NVIDIA Jetson AGX Orin, which is designed for a variety of AI verticals that require high-performance embedded computers at the edge.

Trespasser detection solution in farmlands: the solution involves an AI model that can detect trespassers who climb over a wall or a fence based on videos or images received by cameras installed near the wall or fence. The model is deployed/installed on Aetina’s Mini M1 (AIE-CN11/21 | AIE-CT41). The farmland owners can adopt more than one Mini M1 device and deploy the device(s) near all walls or fences, and manage the device(s) with EdgeEye on Aetina’s AI inference platform AIP-SQ67. In addition to running AI inference tasks, the AIP-SQ67 is able to function as the central computer for its user to monitor the real-time running status of all edge devices with EdgeEye in a web browse.

For more information, visit aetina.com.