01 AI Accelerators Are Becoming Integrated Into Computer on Module Architectures The need for local AI processing across industrial automation, robotics, medical imaging, and edge devices is encouraging the integration of AI accelerators directly into Computer-on-Module architectures. This transition combines CPUs, GPUs, and dedicated NPUs within a single module, reducing reliance on separate accelerator cards and simplifying embedded system designs. In January 2026, congatec introduced Computer-on-Module products based on Intel Core Ultra Series 3 processors with up to 180 TOPS of AI performance, including an NPU delivering up to 50 TOPS, highlighting the expanding AI capabilities embedded within Computer-on-Module platforms.
02 Computer on Modules Are Moving Toward Smaller Form Factors for Space-Constrained Devices The need for compact computing in robotics, portable medical equipment, industrial controls, and edge IoT devices is driving Computer on Module designs toward smaller footprints without removing essential processing and connectivity capabilities. This transition allows module manufacturers to fit computing, memory, and interfaces into space-constrained systems while maintaining standardized integration with carrier boards. Toradex's OSM iMX91, for example, uses a 30 × 30 mm form factor with up to 2 GB LPDDR4 memory and 256 GB eMMC storage, illustrating the move toward highly compact embedded computing modules.
01 Industrial Automation Expansion and Modular Design Drive Computer on Module Adoption Industrial automation expansion creates demand for Computer on Module (COM) platforms used in machine control, robotics, machine vision, and industrial monitoring. In September 2026, the International Federation of Robotics reported that more than 600,000 industrial robots were installed globally in 2025, while the worldwide operational stock reached 5 million robots.Modular hardware design supports Computer on Module adoption by separating application-specific carrier-board development from complex processor, memory, and computing functions. Current COM platforms integrate key computing components into validated modules, allowing developers to reuse the same carrier architecture while changing the computing module for different performance requirements.
01 High Costs and Compatibility Issues Restrain Market Expansion High development and integration costs increase the investment required for module customization, carrier-board design, software integration, and system validation. The resulting cost burden can delay adoption among cost-sensitive OEMs and embedded-system developers, limiting wider deployment of Computer on Module solutions.Hardware and software compatibility issues can create integration difficulties between Computer on Modules, carrier boards, operating systems, drivers, and peripheral interfaces. Additional engineering and testing requirements can extend development timelines and increase costs, which can slow deployment and restrict wider adoption.
01 Medical Imaging and Diagnostic Equipment and Autonomous Mobile Robotics Create New Growth Opportunities for Computer on Modules Medical-device manufacturers, diagnostic equipment companies, and embedded computing providers can use Computer on Modules for imaging, monitoring, and diagnostic systems. Kontron reported EUR 1.61 billion in 2025 revenue and operates across medical technology and connectivity applications, supporting opportunities for application-specific COM integration, software, and system services.Robotics manufacturers, autonomous mobile robot developers, and embedded solution providers can integrate Computer on Modules for navigation, sensor processing, and onboard control. Kontron employs around 7,000 people across more than 20 countries, while its COM portfolio includes technologies such as COM Express and COM-HPC, creating revenue avenues through robotics-focused computing platforms, integration, and software services.
01 Semiconductor and Processor Supply Disruptions Hinder Computer on Module Market Growth Computer on Module (CoM) suppliers depend on processors, memory, power-management ICs, and other specialized semiconductor components that can face allocation or lead-time disruptions. Such shortages can delay module production and make it harder for suppliers to maintain delivery schedules across embedded-system programs.The global semiconductor shortage disrupted embedded computing supply chains and caused extended lead times for processors and other components used in computer modules. These disruptions forced embedded-system manufacturers to redesign or postpone products, demonstrating how component availability can slow CoM market expansion.
June 2026 Advantech expanded its Computer-on-Module portfolio with new COM Express and SMARC modules featuring next-generation processors, enhanced AI computing capabilities, and improved connectivity for industrial automation and edge computing applications.
May 2026 congatec launched new Computer-on-Module solutions based on advanced processor platforms, strengthening its embedded computing portfolio for industrial automation, medical devices, transportation, and AI edge applications.
March 2026 Kontron expanded its embedded computing portfolio with new COM Express modules featuring higher processing performance, extended lifecycle support, and enhanced security features for industrial and IoT applications.
January 2026 Toradex introduced upgraded System-on-Module and Computer-on-Module solutions with improved processing capabilities, connectivity options, and software support for embedded system developers.