Forlinx Embedded is a specialized designer and manufacturer of ARM-based embedded computing solutions. The company’s main product portfolio includes System-on-Modules (SoMs), Single Board Computers (SBCs), and industry-grade PCs. Forlinx maintains deep technical partnerships with top System-on-Chip vendors like NXP, Texas Instruments, Rockchip, and Allwinner to provide reliable hardware foundations for industrial applications.
A key highlight of their current activity is the introduction of the new Pin-to-Pin compatible SoM series—UP4 Family. With this UP4 P2P series, you can directly replace an older one without requiring changes to the baseboard design, wiring, or software. This means products can easily scale across performance tiers, allowing the same carrier board to support different SoMs depending on project requirements.
What are the main areas of activity of the company?
The company’s primary activities center on providing reliable embedded products with a strong emphasis on Linux OS support, hardware security, and long-term industrial availability. These solutions enable faster time-to-market for customers across a diverse range of mission-critical industries, including Industrial Automation, Robotics, Healthcare, Test and Measurement, Transportation, Agriculture, Smart City infrastructure, Renewable Energy, and high-reliability sectors such as Defense and Aerospace Technology.
Forlinx provides the tools, expertise, and software ecosystem to help you build powerful AI-enabled edge solutions. Whether you’re developing innovative LLM-based applications, building next-gen predictive maintenance software, or scaling machine vision products across your industry, our products will be your trusted partner.
What’s the news about new products/services?
Forlinx Embedded has recently expanded its portfolio with several high-performance modules optimized for Machine Learning and Edge AI.
The new FET-MX9596-C SoM is built on the NXP i.MX95 processor, featuring six ARM Cortex-A55 cores and an integrated 6 eTOPS* NPU for AI acceleration alongside 4K video support.
Additionally, the FET-MX8MPx-SMARC module provides 2.3 TOPS of NPU power within a standard SMARC 2.1 form factor, designed to thrive in harsh environments ranging from –40°C to +85°C.
For low power consumption edge AI projects, we have introduced the FET1126B-S module based on Rockchip RV1126B, which offers a 3 TOPS NPU optimized for real-time video analysis and target recognition, backed by a 10-15 year longevity guarantee.
To address growing cybersecurity regulations such as the EU Cyber Resilience Act (CRA), Forlinx presented its CRA-ready embedded platform, the FCU2601. With dual IEC 62443 certifications, the platform helps device manufacturers accelerate compliance with global cybersecurity regulations while ensuring robust industrial-grade security.
Having gone through the entire certification process ourselves, we are deeply familiar with the complex documentation and compliance workflow. This experience allows us to share proven practices with customers and help simplify the path to market, especially for products targeting the European market.
What are the ranges of products/services?
Our product range encompasses a comprehensive ecosystem of hardware based on NXP, Texas Instruments, Rockchip, and Allwinner architectures, including system on modules, independent AI acceleration cards, and embedded AI edge gateways.
Beyond hardware, we have powerful in-house R&D team offer optimized software development experience including free and prompt support from technical support team, comprehensive hardware and software guides and documentation, carrier board design and reference, design schematics review etc.
Forlinx provides custom solutions tailored to specific industry requirements, encompassing carrier board and SoM customization, kernel optimization, driver development, and system porting
What is the state of the market where you are currently active?
The embedded market is shifting from simple connectivity to Edge AI, driven by the efficiency of ARM architecture. ARM’s superior power-to-performance ratio makes it the standard for industrial intelligence, offering a mature ecosystem for both real-time and high-performance applications.
The most significant trend is the transition toward System-on-Modules (SoM). Instead of complex “chip-down” designs, engineers are adopting modular SoMs to slash development time, reduce risk and ensure Longevity.
In 2026, success in this market is defined by supply chain resilience and the ability to provide 10–15 years of guaranteed availability for these modular solutions.
What can you tell us about market trends?
While demand for traditional industrial control remains relatively stable, the rapid development and popularization of AI have led to a significant increase in customer projects related to intelligence and edge computing.
Simultaneously, the global memory chip market continues to face supply constraints, resulting in extended lead times for certain components and noticeable price volatility. These conditions present ongoing challenges for equipment manufacturers regarding production planning and effective cost control.
What are the most innovative products/services marketed?
The most innovative offering currently in our lineup is the FAI-ARA240-M Edge AI Accelerator, which is specifically designed for real-time inference of Generative AI and Large Language Models (LLMs).
This M.2 (M-Key) interface card delivers up to 40 eTOPS of performance with 16GB of LPDDR4 memory and features a programmable DNPU that supports Transformer, CNN, and multimodal models.
It is fully compatible with mainstream frameworks like TensorFlow and PyTorch and includes integrated Secure Boot and Root-of-Trust processing to meet the security requirements of modern embedded systems.
What estimations do you have for the beginning of 2026?
Heading into early 2026, we anticipate a continued surge in demand for intelligent edge devices that require increased local computational power to handle high-performance tasks. We expect customers to place an even higher priority on supply chain stability and long-term supply guarantees.
To address this, Forlinx Embedded will focus on delivering a wide range of industrial-grade core boards with integrated AI capabilities and providing comprehensive ODM customization services. Our goal is to offer an integrated partnership—from hardware selection to system support—allowing our customers to minimize their investment in underlying hardware development and focus their resources on proprietary software applications.
*eTOPS=equivalent TOPS
