Innodisk Unveils Comprehensive Qualcomm Dragonwing Edge AI Series at Computex 2026
Innodisk, a prominent player in the global AI solutions market, made significant waves at Computex 2026 by unveiling its expansive collaboration with Qualcomm Technologies, Inc. This introduction showcased the full range of its 'AI on Dragonwing' computing series that was initially launched earlier in the year. The technology has rapidly evolved, now encompassing four generations of Qualcomm Dragonwing processors, which promise to deliver remarkable performance scalability ranging from 20 to an impressive 700 TOPS (Tera Operations Per Second) tailored to diverse edge AI applications.
The ongoing transition towards edge AI in domains such as factory automation, smart cities, and defect inspection has accentuated the need for energy-efficient inference capabilities. In particular, for battery-operated Autonomous Mobile Robots (AMRs) and Automated Guided Vehicles (AGVs), minimizing energy consumption is essential to maintain their operational longevity while also reducing thermal costs. This makes the new high-performance and low-power architecture a prime candidate for next-generation implementations.
Qualcomm's ambition with the Dragonwing architecture is to provide outstanding power efficiency combined with multifunctional computing capabilities aimed at the industrial sector. The extensive lineup includes the Qualcomm Dragonwing™ models IQ8, IQ9, IQ10, and the IQ-X Series, each offering long product lifespan support that extends through 2036 or beyond. This diverse array of processors is designed to empower enterprises to optimize their computing infrastructures effectively.
To leverage the potential of this state-of-the-art technology, Innodisk has introduced the 'AI on Dragonwing' hardware assortment, which includes modules, starter kits, and ready-to-deploy systems. Coupled with the IQ Studio software toolkit, which simplifies the transition from evaluation to deployment—complete with Board Support Package (BSP), benchmarking tools, and model optimization—Innodisk positions itself to facilitate seamless integration of these advanced features into various applications.
The lineup presented at Computex 2026 showcases four distinct Dragonwing platforms, each tailored to unique application profiles.
At the Innodisk booth during Computex 2026, attendees were treated to two live demonstrations validating the practicality of these systems. The first demonstration featured the EXMP-Q911 module powered by the Dragonwing IQ9 processor, executing real-time defect inspections using CNN algorithms alongside DMS and vision-linked Personal Protective Equipment (PPE) detection within a stringent total system power output of just 30W. Significantly, the system achieved an ultra-low average inference latency of 4.4 milliseconds, marking a 2.6x performance advantage over typical existing architectures.
The second demonstration illustrated the capabilities of vision-powered autonomous robotic navigation utilizing the EXFC-Q911 fanless system, which is specifically engineered for AMRs and AGVs. This setup managed a high-throughput 8-channel GMSL2 camera system and 3D depth sensors, providing comprehensive spatial awareness and real-time obstacle avoidance, all while operating at an efficient power draw of just 32W.
In summary, Innodisk's commitment to pushing the boundaries of AI technology with the support of Qualcomm’s Dragonwing processors signifies an important milestone for the industry, focusing on performance, efficiency, and future technologies that are ready to meet the demands of tomorrow’s automated environments.
Industry-Leading Architecture for Diverse Needs
The ongoing transition towards edge AI in domains such as factory automation, smart cities, and defect inspection has accentuated the need for energy-efficient inference capabilities. In particular, for battery-operated Autonomous Mobile Robots (AMRs) and Automated Guided Vehicles (AGVs), minimizing energy consumption is essential to maintain their operational longevity while also reducing thermal costs. This makes the new high-performance and low-power architecture a prime candidate for next-generation implementations.
Qualcomm's ambition with the Dragonwing architecture is to provide outstanding power efficiency combined with multifunctional computing capabilities aimed at the industrial sector. The extensive lineup includes the Qualcomm Dragonwing™ models IQ8, IQ9, IQ10, and the IQ-X Series, each offering long product lifespan support that extends through 2036 or beyond. This diverse array of processors is designed to empower enterprises to optimize their computing infrastructures effectively.
To leverage the potential of this state-of-the-art technology, Innodisk has introduced the 'AI on Dragonwing' hardware assortment, which includes modules, starter kits, and ready-to-deploy systems. Coupled with the IQ Studio software toolkit, which simplifies the transition from evaluation to deployment—complete with Board Support Package (BSP), benchmarking tools, and model optimization—Innodisk positions itself to facilitate seamless integration of these advanced features into various applications.
Scalable Performance Across Diverse Applications
The lineup presented at Computex 2026 showcases four distinct Dragonwing platforms, each tailored to unique application profiles.
- - Dragonwing IQ8 Series: This series serves as the power source for the EXMP-Q801 COM-HPC Mini module, offering capabilities of up to 40 TOPS (Dense) and is engineered for wide-temperature applications including vision inference and sensor fusion.
- - Dragonwing IQ9 Series: The IQ9 Series scales performance capabilities up to 100 TOPS (Dense) through its dual Neural Processing Unit (NPU) and powers various components such as the EXMP-Q911 module, EXEC-Q911 starter kit, and the APEX-A100 fanless edge AI system. This setup is vital for applications like industrial inspections and Driver Monitoring Systems (DMS).
- - Dragonwing IQ10 Series: With the EXMR-QA01 module, this series extends capabilities markedly to a staggering 350 TOPS (Dense). It boasts an 18-core Oryon™ CPU and supports up to 20 MIPI cameras at a resolution of 16 MP, along with an integrated Safety Island (SAIL) module that is custom-designed for robust AMR navigation, humanoid robotics, and other heavy-duty edge applications.
- - Dragonwing IQ-X Series: This series distinguishes itself with its native Windows compatibility. The EXDU-QX11 edge AI system, leveraging the Dragonwing IQ-X processor, provides up to 45 TOPS and supports dual operating systems—Windows 11 IoT Enterprise LTSC and Ubuntu 26 LTS—making it ideal for deployment in programmable logic controllers (PLCs), Human-Machine Interfaces (HMIs), and edge controllers, equipped with various connectivity features.
Live Demos Highlighting Deployment Readiness
At the Innodisk booth during Computex 2026, attendees were treated to two live demonstrations validating the practicality of these systems. The first demonstration featured the EXMP-Q911 module powered by the Dragonwing IQ9 processor, executing real-time defect inspections using CNN algorithms alongside DMS and vision-linked Personal Protective Equipment (PPE) detection within a stringent total system power output of just 30W. Significantly, the system achieved an ultra-low average inference latency of 4.4 milliseconds, marking a 2.6x performance advantage over typical existing architectures.
The second demonstration illustrated the capabilities of vision-powered autonomous robotic navigation utilizing the EXFC-Q911 fanless system, which is specifically engineered for AMRs and AGVs. This setup managed a high-throughput 8-channel GMSL2 camera system and 3D depth sensors, providing comprehensive spatial awareness and real-time obstacle avoidance, all while operating at an efficient power draw of just 32W.
In summary, Innodisk's commitment to pushing the boundaries of AI technology with the support of Qualcomm’s Dragonwing processors signifies an important milestone for the industry, focusing on performance, efficiency, and future technologies that are ready to meet the demands of tomorrow’s automated environments.
Topics Consumer Technology）
【About Using Articles】
You can freely use the title and article content by linking to the page where the article is posted.
※ Images cannot be used.
【About Links】
Links are free to use.