MediaTek and Google bringing Gemini Nano to more smartphones
Google‘s Gemini Nano is an on-device AI model and only a few devices could support this as it requires a lot of computing power and memory. Now, the company is planning to expand its availability to more devices.
MediaTek has announced that some of its chips are now optimized for Gemini Nano. According to the information, MediaTek and Google have collaborated to integrate and optimize Gemini Nano on the MediaTek Dimensity 9300 and 8300 chipsets, which are designed for high-end and mid-range smartphones, respectively.
Follow our socials → Google News | Telegram | X/Twitter | Facebook | WhatsApp
The press release states that this optimization includes using MediaTek’s NeuroPilot toolkit and porting Gemini Nano to MediaTek’s APU. This will enable Gemini Nano to run more effectively and efficiently on MediaTek-powered devices, without compromising battery life or performance.
MediaTek also says that it will soon release an APK for developers and OEMs to help them deploy Gemini Nano applications on their devices. Additionally, MediaTek will showcase its on-device generative AI capabilities using Llama 2 LLM at the Mobile World Congress 2024 next week.
With this partnership, MediaTek and Google aim to bring on-device generative AI to more users and create new possibilities for smartphone applications and experiences.