ADLINKジャパン List of News Articles

The event hosted by AMD is a technology conference showcasing the latest products and solutions featuring AMD technology. AMD is holding this event with the aim of accelerating "trust" in AI together with its business partners and end users. ADLINK plans to exhibit its latest AMD-powered computer-on-module. ●Type 6 Compact module "cExpress-R8"

● Based on the NXP i.MX 95, it achieves AI performance of up to 2 TOPS. Compliant with SGeT SMARC 2.2 and OSM 1.2, it provides scalable and efficient edge intelligence. ● OSM-IMX95: A robust 45mm x 45mm solderable module adopting ADLINK's single-sided PCB design. It offers AI-driven performance for compact, power-constrained edge devices. ● SMARC LEC-IMX95: An 82 mm x 50 mm SGeT SMARC 2.2 module. Equipped with 10 GbE, secure boot, and industrial reliability, it enables intelligent IoT and automation. ● Providing customized high-performance edge solutions for various industries through diverse options and flexible ODM customization.

NVIDIA, a partner company of ADLINK Japan, will hold the "Latest Technology Session by NVIDIA Partners" at the 2025 Japan IT Week [Autumn], where ADLINK Japan's Kiuchi will present on the theme "Introduction to Embedded AI Solutions for TCO Reduction." This is an open seminar that is free and does not require prior registration, so please join us.

The event hosted by AMD is a technology conference that showcases the latest products and solutions featuring AMD technology. Through presentations by experts and interactive demonstrations, AMD's initiatives that lead cutting-edge technological trends in a wide range of fields such as manufacturing, AI, robotics, healthcare, automotive, aerospace, and defense will be introduced. There will also be opportunities for direct interaction with industry experts. The content is designed to provide an optimal opportunity for those involved in design and development to deepen their understanding of efficient application development using AMD's platforms and gain new insights. ADLINK is scheduled to showcase the latest AMD-based computer-on-modules, the "Type 6 Compact Module cExpress-R8" and the "Type 7 Basic Module Express-VR7." We look forward to your visit.

● AI-enabled high performance: Equipped with an integrated NPU and optionally compatible with NVIDIA GPUs, enabling scalable edge AI workloads. ● Modular I/O and flexible scalability: Supports function modules, spare signals, and extensive I/O customization to adapt to diverse application needs. ● Open-frame industrial design: Robust structure and wide operating temperature range allow for seamless integration into kiosks, machines, terminals, and more. ● High-quality display: Features industrial-grade AUO panels with ADLINK's unique customizations, including optical bonding, high brightness, anti-reflection, and UV protection.

●Equipped with AMD Ryzen Embedded 8000 series processors, achieving overwhelming performance. ●Accelerates multitasking and immersive visual experiences with 8-core, 16-thread processing powered by AMD RDNA 3 graphics. ●Achieves up to 40 TOPS of AI inference performance with Zen 4, RDNA 3, and XDNA architectures, enhancing on-device intelligence. ●Supports ECC-enabled DDR5 memory up to 96GB, ensuring reliable data integrity and system stability. ●Compatible with a variety of multi-display configurations (DP, eDP, HDMI, LVDS, VGA), flexibly accommodating complex visual setups.

●Wide voltage input: Supports 12-28V DC input, ensuring stable operation even in environments with unstable power. ●Balanced performance: Supports Intel Core CPUs with a maximum of 65W using PCIe 4.0, DDR4, and dual 2.5GbE. ●Ideal design for industrial applications: The Mini-ITX form factor and rich expandability enable seamless integration into embedded systems.

● Next-generation DLAP expandable edge AI platform: Built to meet the rapidly growing demand for edge AI solutions with flexible scalability and robust design. ● High-performance AI computing: Achieves scalable AI processing capabilities of up to 91.1 TFLOPS with support for NVIDIA RTX 6000 ADA GPU, enabling real-time decision-making at the edge. ● Versatile industrial applications: Optimized for mission-critical uses in manufacturing, healthcare, logistics, and smart cities, enhancing operational efficiency and intelligence.

●3.5 "Expanding the SBC Portfolio: ADLINK has announced the SBC35-MTL and SBC35-ASL, targeting both high-performance AI and robust embedded edge applications. ●AI Compatibility and Robust Design: The SBC35-MTL is equipped with an Intel Core Ultra featuring an NPU for real-time inference, while the SBC35-ASL is powered by the Intel Atom x7000RE, providing a wide temperature range and fanless operation. ●Flexible I/O with SBC-FM Modules: Both models support ADLINK's SBC-FM expansion modules, enabling customized I/O to accelerate deployment across various industries.

● High-Performance Computing: The EMP-520 series, equipped with the 14th generation Intel Core processor, delivers powerful processing capabilities, ensuring seamless multitasking and efficient management of heavy workloads. ● Excellent Visual Experience: With EDID emulation functionality, it supports simultaneous output of 4K video, providing vibrant and consistent displays ideal for dynamic digital signage environments. ● Reliable Design: The EMP-520 series is designed to withstand industrial environments, guaranteeing reliable operation 24/7, 365 days a year.

- Equipped with the Intel Core Ultra architecture featuring up to 14 CPU cores, 8 Xe GPU cores, and an integrated NPU for high-performance AI acceleration, offering both ultra power and energy efficiency. - 64GB of LPDDR5x memory is directly soldered onto the board, supporting maximum performance in a 95 x 70mm form factor, with an operating temperature range of -40°C to 85°C. - Integrated with up to 16 PCIe lanes, two SATA interfaces, two 2.5GbE Ethernet ports, and multiple DDI/USB4 and USB 3.0/2.0 interfaces for abundant I/O options.

This year, Japan's largest IT exhibition, "Japan IT Week," where cutting-edge IT solutions converge, will be renewed and held at Tokyo Big Sight. Our company will be exhibiting at the notable IoT and Edge Computing EXPO. At this exhibition, we plan to showcase in-vehicle camera solutions, real-time object detection x monocular depth estimation, edge AI, GPU solutions, edge computing, industrial panels and monitors, among others. We will also introduce the latest technologies, including ready-to-use SMARC development kits and portable Pocket AI. ★ This time, we will showcase collaborative demos with five companies ★ 【TIER IV】 In-vehicle camera solutions 【Kansai University Graduate School of Comprehensive Informatics, Tanaka Laboratory】 Real-time object detection x monocular depth estimation 【ASK】 Portable external GPU 【Allxon】 Remote management of edge AI devices 【Macnica】 High-speed AI processing by Hailo

●ADLINK has partnered with LIPS to announce an AMR 3D x AI solution using NVIDIA Isaac Perceptor. This solution is designed for autonomous mobile robots with enhanced 3D vision, a wide FOV, and high resolution, surpassing LiDAR. ●This solution includes the LIPSAMR Perception DevKit equipped with the DLAP-411-Orin and LIPSedge 3D camera, providing high-precision 3D recognition for smart manufacturing and warehouse logistics. ●The ADLINK DLAP-411-Orin platform and the LIPSAMR Perception DevKit equipped with NVIDIA Jetson AGX Orin deliver 275 TOPS of AI performance and support multi-camera synchronization for complex multi-vision applications.

The MXE-230, equipped with the Hailo-8 AI accelerator, can achieve 26 TOPS of edge AI computing capability. The low-power MXE-230, which integrates the Hailo-8 AI accelerator, is particularly suitable for scenarios that require AI processing capabilities while being sensitive to power consumption. Hailo is a global leading provider of high-performance accelerators specifically designed to run complex AI models on edge devices.

●Supports MediaTek Genio 510, achieving intensive workloads and low power consumption ●Offers up to 8GB LPDDR4 RAM, up to 128GB eMMC, 4K graphics support, 30MP ISP camera, and a rich variety of I/O options ●Operates in temperatures from 40℃ to 85℃, enduring long-term mission-critical use with a 10-year lifecycle ●Size-L module compliant with OSM R1.1

- Breakthrough in Generative AI Performance: The DLAP Supreme series equipped with Phison's aiDAPTIV+ technology overcomes memory limitations and performance bottlenecks of edge devices, achieving 8 times the inference speed and 4 times the token length, supporting the training of generative language models. - Core Value of aiDAPTIV+ Technology: Phison's aiDAPTIV+ technology plays a crucial role in efficiently running large language models on devices with different memory configurations, enabling expansion to more application scenarios. - Close Collaboration Driving Industry Progress: ADLINK and Phison have been collaborating for over 10 years. The release of the DLAP Supreme series further promotes the adoption of AI in smart manufacturing, smart cities, and other industries.