specification AI CORE The first embedded ultra-compact Artificial Intelligence processing card for on the edge computing AI Core is a mini-PCI Express module that enables Artificial Intelligence on AI Core enables rapid prototyping, validation and field deployment of the Edge. Deep Neural Network (DNN) inference applications at the edge. Its Thanks to its form factor it can seamless integrated in most of the industrial low-power VPU architecture enables an entirely new segment of AI computing edge devices. applications that are not reliant on a connection to the cloud. It is powered by low power high performance Intel Movidius Myriad 2 AI Core combined with Intel Movidius Neural Compute SDK allows deep 2450 VPU with 512MD DDR that can be found in millions of smart security learning developers to profile, tune, and deploy Convolutional Neural cameras, gesture controlled drones, industrial machine vision equipment, Network (CNN) on low-power applications requiring real-time inferencing. and more. For mobile and connected devices where battery life is critical, Myriad 2 An ultra-low power AI Core - Specifications provides a way to combine advanced vision applications in a low power design profile. Intel s Movidius Myriad 2 VPU delivers vision capabilities to classes of devices previously unable to perform such demanding vision tasks. SOC Intel Movidius Myriad 2 2450 Supported Frameworks Featuring 12 VLIW programmable SHAVE cores, dedicated vision accelerators Unique design for TensorFlow, Cae and 2 CPUS, all connected by an intelligent memory fabric, Intel s Movidius vision and AI Myriad 2 VPU is a fully functional vision SoC designed for high performance workloads Form factor at ultra-low power. The highly parallel design and optimizations for sparse Mini PCI-Express data structures make Myriad VPUs well suited to deep neural network applications and other modern vision workloads. Dimensions xxx x xxxx mm Minimum system requirements The flexibility for developers to implement differentiated and proprietary 12 programmable x86 64 computer running Ubuntu 16.04 applications is fundamental to Intel Movidius Myriad 2 VPUs. Our SHAVE cores optimized software libraries give device manufacturers the ability to run Available mPCI-E slot custom and proprietary operations on the 12 high performance SHAVE cores. 1GB RAM 4GB free storage space To conserve space inside mobile, wearable, and embedded devices, Intel s A small-area footprint Movidius Myriad 2 was designed with a very small footprint that can easily be integrated into existing products. The Intel Movidius Myriad 2 VPU architecture comprises a complete set of Additional Chip Details interfaces, a set of enhanced imaging/vision accelerators, a group of 12 specialized vector VLIW processors called SHAVEs, and an intelligent memory fabric that pulls together the heterogeneous resources to enable power efficient processing. www.up-board.org /www.up-shop.org / www.up-community.orgEstablished in 1992, AAEON is one of the leading designers and UP is a brand of AAEON Europe, an associate company of ASUS group, manufacturers of advanced industrial and embedded computing focused on Embedded, Internet Of Things, Industry 4.0. UPs mission is platforms today. Committed to innovative engineering, AAEON provides to spread the intelligence in all the world around us. UP bridges the gap integrated solutions, hardware and services for premier OEM/ODMs and between the world of prototypes and the world of high-grade mass-pro- system integrators worldwide. duced embeddedsystems solutions.UP gives affordable and easily accessible and support of all the latest technologies to embeded With a continuous pursuit of innovation and excellence, AAEON became industry. a member of the ASUS group in 2011, further strengthening its leadership fueled by advanced technology from ASUS and leveraging For more information about AAEON, please visit: resources within the group. AAEON is posed to offer more diversified www.up-shop.org embedded products and solutions at higher quality standards to meet world-class design and manufacturing demands in the years to come. For more information about AAEON, please visit: www.aaeon.com Artifical Intelligence on the edge www.up-board.org/ www.up-shop.org /www.up-community.org