On April 3rd, reports surfaced indicating that Google has disabled its latest artificial intelligence model, Gemini Nano, on the Pixel 8 smartphone due to "hardware limitations".
According to Seang Chau, Vice President of Google's Device and Services Software, some AI models within Gemini Nano require constant presence in the device's memory for on-the-fly access. However, the Pixel 8, equipped with only 8GB of memory, risks compromising user experience if the Gemini Nano feature is enabled.
In contrast, the Pixel 8 Pro, with 12GB of memory, is better equipped to leverage the performance of AI models.
From this explanation, it appears that the issue with Pixel 8 isn't the inability to utilize Gemini Nano but rather the significant memory footprint it imposes, impacting user experience.
Comparatively, the 8GB memory variant of the Samsung Galaxy S24 series can run Gemini Nano models without such restrictions.
Originally positioned for on-device execution on mobile devices, Google's Gemini Nano faces notable drawbacks in pure on-device execution, including limited computational resources and the consumption of device memory.
