You are here: Events / RUBIK Pi AI Events
Language: English
Date: 11 Sep 2025
Time: 09:00 - 17:00
Entry fee: 299€* (net + VAT) participation incl. Rubik Pi, touch display, camera, power supply and cable (recommended to book min. 1x per company) // 49€* (net + VAT) event participation only
We would like to invite you to join our interactive RUBIK Pi AI Events, which will take place from September 09, 2025, in four European cities. Together with Consult Red, Edge Impulse, and Thundercomm, Inc., we host four RUBIK Pi AI Events in Munich, Eindhoven, Copenhagen, and London, offering you a unique opportunity to explore the versatile application options of the first single-board computer based on a Qualcomm Technologies SoC, the Dragonwing™ QCS6490, and test its AI capabilities directly in person.
Come join us this fall and learn how to utilize RUBIK Pi’s AI capabilities for developing the next generation of intelligent edge devices!
Rubik Pi AI Event Schedule
| Time | Agenda |
|---|---|
| 9:00 – 9:30 | Registration |
| 9:30 – 10:00 | Welcome (Atlantik Elektronik) |
| 10:00 – 11:00 | Qualcomm Technologies Portfolio (Atlantik Elektronik) |
| 11:00 – 11:30 | Qualcomm Linux Software Stack | Developer First (Atlantik Elektronik) |
| 11:30 – 12:30 | Lunch and Demos (Showroom) |
| 12:30 – 13:30 | Rubik Pi Introduction (Thundercomm) |
| 13:30 – 14:00 | Multimedia SW and Architectur: Display, Cameras, Audio (Thundercomm) |
| 14:00 – 16:00 | Edge Impulse Training on how to use, deploy and train models on Rubik Pi (Edge Impulse) |
| 16:00 – 17:00 | Real Results: How We Built a Visual Anomaly Detection Solution (Consult Red) |
9:00 - 10:00 Registration and Welcome
With almost 50 years of experience in bringing the latest technology to life, Atlantik Elektronik is the expert for new technological development. As a partner of Thundercomm, Inc. and authorized distributor for Qualcomm Technologies, Inc., we can provide interesting insights into the innovative technology behind RUBIK Pi and its potential for more successful product development with the help of an entire technology ecosystem. Whether you are planning a chip on board design, want to use a system on module, receive a custom PCBA or even a turnkey solution, RUBIK Pi can be the best option for your project.
With Consult Red’s and Edge Impulse’s experience in integrating AI models on RUBIK Pi3, their expertise will enrich the more interactive part of the event and provide you with comprehensive insights into the use, deployment, and training of AI models on RUBIK Pi.
10:00 - 11:00 Qualcomm Technologies Portfolio (Atlantik Elektronik)
Our Atlantik team will introduce the latest developments of the Qualcomm Technologies Portfolio:
11:00-11:30 Qualcomm Linux Software Stack | Developer First (Atlantik Elektronik)
With the new comprehensive Linux software solution, Qualcomm Technologies offers a highly advanced development stack optimized for the specific requirements of IoT development projects. The unified Linux distribution supports various System-on-Chip (SoC) platforms and includes essential components such as UEFI boot loader, Long-Term Support (LTS) kernel, Yocto and specially selected drivers, enabling consistent and smooth developer experience regardless of the target system.
11:30 - 12:30 Lunch and Demos (Showroom)
12:30 - 13:30 Rubik Pi Introduction (Thundercomm)
The Thundercomm RUBIK Pi 3 lightweight development board is based on the Qualcomm Dragonwing™ QCS6490, making it the first Pi built on Qualcomm AI platforms for developers.
In its ultra-compact and lightweight format, measuring only 100mm x 75mm, the microcomputer integrates the high-performing Qualcomm Dragonwing™ QCS6490 processor, which delivers 12.5 TOPS of on-device AI inference performance and supports small language models with up to 1.8 billion parameters. It also features memory, storage, and a variety of interfaces and functions, such as i.e. the support of USB Type-A (1x 2.0, 2x 3.0), USB 3.1 Gen 1 Type-C with DP (4K@60), Camera, HDMI OUT (4K@30), a 1000M Ethernet port, and more.
RUBIK Pi 3 supports multiple operating systems such as Android/ Linux/ LU with SDKs from Thundercomm to provide developers with improved experience. Thanks to Qualcomm Dragonwing™ QCS6490’s excellent AI performance at 12 TOPS and compatibility with Qualcomm AI Hub developers can directly benefit from the impact of AI evolution.
13:30 - 14:00 Multimedia SW and Architecture: Display, Kameras, Audio (Thundercomm)
This session provides an overview of the multimedia software stack and architecture on the Thundercomm Rubik Pi, running Qualcomm Technologies’ Yocto-based Linux OS. It will cover key components including display interfaces, camera integration pipelines, and audio subsystems, highlighting how these are managed within the Qualcomm multimedia framework. Attendees will gain insights into GStreamer-based camera and video processing, display handling through DRM/KMS, and ALSA/Audio DSP integration for robust audio support. The training will also touch on multimedia HAL layers, enabling efficient interaction with hardware acceleration features of the Snapdragon SoC.
14:00 - 16:00 Edge Impulse Training on how to use, deploy and train models on Rubik Pi (Edge Impulse)
This technical training session will provide a practical understanding of the Edge Impulse platform and how it fits into the broader Qualcomm Technologies ecosystem for edge AI development.
We will cover the entire machine learning development pipeline, starting from data collection and preprocessing via model training, optimization, and validation to deployment. Key features such as signal processing, anomaly detection, transfer learning, quantization, and the EON compiler will be explored through real-world use cases across audio, vision, and sensor data.
A hands-on lab session will guide participants through deploying a trained model onto the Thundercomm RubikPi 3 with live demos to showcase real-time inference at the edge.
By the end of the session, attendees will be equipped to support and advocate for Edge Impulse solutions across a range of verticals and Qualcomm platforms.
16:00 - 17:00 Real Results: How We Built a Visual Anomaly Detection Solution (Consult Red)
What does it really take to bring an Edge AI solution to life? We will present behind-the-scenes insights into the process of designing and building a production defect detection demonstrator from proof-of-concept to practical use.
We will explain the design and development of a complete Edge AI prototype, built to detect faults in pharmaceutical blister packs on a conveyor belt using real-time video analysis. Participants will learn how embedded hardware, machine learning models (managed with Edge Impulse tooling), and intelligent image processing were combined to efficiently deliver industrial-grade results on a Thundercomm RUBIK Pi3 development board (powered by Qualcomm Dragonwing™ QCS6490).
Qualcomm-branded products are products of Qualcomm Technologies, Inc. and/or its subsidiaries. The Linux Foundation and The Linux Foundation logo design are registered trademarks of The Linux Foundation. Linux is a registered trademark of Linus Torvalds.
