TLDR:
Google announced that only the Pixel 8 Pro will be able to run its latest AI model, “Google Gemini,” due to RAM limitations. The Pixel 8 will get the model as a developer option in June. Google is keeping some AI models “RAM-resident,” leading to concerns about system memory trade-offs. The baseline for RAM in phones may need to increase in the future.
Key Elements:
- Google announced that only the Pixel 8 Pro will be able to run its latest AI model, “Google Gemini,” due to RAM limitations.
- The Pixel 8 will get the model as a developer option in June, while the Pixel 8 Pro will have it as a user-facing feature.
- Google is keeping some AI models “RAM-resident,” leading to concerns about system memory trade-offs and potential need for increased RAM in future phones.
In early March, Google announced that only the Pixel 8 Pro will be able to run its latest AI model, “Google Gemini,” due to hardware limitations. However, in a recent update on the Pixel Phone Help forum, Google stated that the Pixel 8 will also receive the model as a developer option in the next quarterly Android release. This decision was explained by Seang Chau, VP of devices and services software at Google, who highlighted the differences in RAM between the two devices as a key factor.
Chau mentioned that AI models like Gemini Nano need to be “RAM-resident” to ensure they are available at all times, leading to potential trade-offs in system memory. This has raised concerns about the need for increased RAM in future Android phones, especially with the increasing demand for AI-powered features.
While Google is pushing AI models, such as Gemini Nano, on smartphones like the Pixel 8 Pro and the Galaxy S24 lineup, users may question the necessity of these features and their impact on system performance. It remains to be seen how the inclusion of AI models in smartphones will affect user experience and system resources in the long run.