HONOR, the global leader in smart technologies, today announced at Snapdragon Summit 2023 that the upcoming HONOR Magic6 premium smartphone series is powered by the Snapdragon 8 Gen 3 mobile platform and showcased the power of the device's own Large Language Model (LLM). with a parameter size of seven billion.
A collaboration between HONOR and Qualcomm Technologies created an on-device LLM and a human-centered ecosystem that enables seamless interactions across multiple devices. At the event, HONOR also introduced several enhanced MagicRing connectivity features.
Hunt for dinosaur bones
“At HONOR, we are committed to pushing the boundaries of innovation and creating a smarter, more intuitive user experience,” said George Zhao, CEO of HONOR. “We are leveraging the power of artificial intelligence to transform the way people interact with technology, making it more seamless, personalized and impactful in our lives, without compromising user privacy. We are pleased to collaborate with Qualcomm Technologies as the company shares our vision of striving for excellence with people at the heart. We will continue to innovate and work with partners in open collaboration to shape a future where AI supports and enriches the lives of consumers around the world.”
“We look forward to working with HONOR to advance technological advancements with cutting-edge technologies,” added Alex Katouzian, Senior Vice President and General Manager, Mobile, Compute and XR Devices at Qualcomm Technologies. “By collaborating with HONOR, we are revolutionizing the way people live, work and connect, creating transformative experiences that open up new possibilities in the ever-evolving world of AI.”
Unlike cloud LLMs that are trained on publicly available datasets, HONOR's on-device LLM relies on the usage habits of the individual device user and offers personalized services according to their preferences. Since the user's data actually remains on the device with an on-device LLM, consumers are exposed to significantly fewer data protection risks with an optimized user experience.
To develop an on-device LLM that meets the needs of the average consumer, HONOR and Qualcomm Technologies collaborated in three key areas: performance, efficiency and user privacy. The joint development resulted in optimizations that allow the mobile platform to effectively leverage the benefits of LLM without consuming too much power, ensuring that data on the device remains safe and secure.
With the Snapdragon 8 Gen 3 mobile platform, the HONOR Magic6 series is powered by an on-device LLM with seven billion parameters. HONOR demonstrated the potential of its AI capabilities during its presentation as part of a demo on creating videos using artificial intelligence. With commands to the YOYO smart voice assistant, users can easily create short videos from footage stored on their devices. Templates, themes and music of the generated videos can be customized and personalized with additional prompts.
At the Snapdragon Summit 2023, HONOR also presented eye control: the device tracks the user's eye movements and allows multimodal interaction with the functions of the so-called MagicRing: a networked camera, input and screen sharing on multiple devices. Using the latest Snapdragon Seamless technology, MagicRing enables seamless data exchange between devices that is simultaneously more stable and more energy efficient than ever before.
Going forward, HONOR will continue to openly collaborate with global partners while cultivating an ecosystem of on-device AI technology. With its innovations, the company contributes to a smarter future that continues to put people at the center.
* = Affiliate link/advertisementAs an Amazon Associate, we earn money from qualifying purchases. If you make a purchase through one of the links, we receive a portion of the purchase price as commission. This doesn't cost you anything extra, but it helps support our website. Last updated on September 19.04.2026, XNUMX / Affiliate links / Images from the Amazon Product Advertising API. Images from Amazon PA API.
