With its 8GB of RAM, the Pixel 9a can only power a less powerful AI model than other Pixel 9s. It therefore lacks several interesting features that are available on Google's higher-end smartphones.
Google launched the Pixel 9a, its new mid-range smartphone, yesterday. The announcement was marked by a small hiccup, however, as the Pixel 9a is still not available for pre-order. The manufacturer was forced to revise its availability schedule after discovering a component issue during the device's production. This would only affect a limited number of phones, but Google prefers to take the time to carry out proper quality control before launching it.
While waiting to test the smartphone, we already have some bad news. While the Pixel 9a is equipped with a Tensor G4 chip, like the other Pixel 9 models, it only has 8 GB of RAM. For comparison, the Pixel 9 has 12 GB of RAM, and the Pixel 9 Pro and Pixel 9 Pro XL have 16 GB.
The Pixel 9a's AI model is less powerful
However, the amount of RAM on a smartphone has never seemed so important as it is in recent times. While Android manufacturers could be criticized for offering a lot of RAM on their devices for marketing purposes only or to compensate for a lack of optimization, RAM has become essential with the emergence of AI, which requires enormous resources to function.
As pointed out by Ars Technica, the Pixel 9a's 8GB of RAM doesn't allow it to access the Gemini Nano XS model, which powers all other phones in the Pixel 9 series. The Pixel 9a has to make do with a new, smaller, and therefore less powerful model, which Google has dubbed Gemini Nano 1.0 XXS.
The first major difference between Gemini Nano 1.0 XXS and Gemini Nano XS is that the latter runs constantly in the background, ready to intervene as soon as the user calls upon its capabilities. The Pixel 9a's local model, on the other hand, only becomes active when it is called upon. This means it will likely be a little slower to generate its results, as there will be a slight lag time for it to wake up from sleep.
A lack of transparency from Google?
But above all, Gemini Nano 1.0 XXS is not compatible with many of Google's most advanced AI features. This model is not multimodal: it only supports text, not images or audio. For example, it does not work with the Pixel Screenshots app. This app, launched at the same time as the Pixel 9, allows you to create a database of your screenshots. AI processing records useful details from each image so that the user can easily find information without having to tediously search through their gallery.
The Call Notes feature is also missing. This generates a summary of your phone conversations, a very useful option in certain situations to save time or not forget important information. Call Notes isn't currently available in French, but we know the Pixel 9a won't get it when it does.
Generally speaking, you can expect most of Google's AI features involving image or audio to be unsupported by the Pixel 9a. There are a few exceptions, though. Generating a summary of an audio recording, for example, works, as the app creates a transcript without using Gemini, which then takes care of providing a summary from the transcript in text form.
It's unfortunate that Google hasn't been entirely transparent about the Pixel 9a's lack of AI features. The smartphone's product page doesn't indicate that it has fewer options of this type than other Pixel 9 devices. It's also worth remembering that the Pixel 9a doesn't offer free access to Gemini Advanced. While the Pixel 9 Pro offers a one-year subscription to this service, the Pixel 9a costs €21.99 per month, with the first month's trial free.
Source: Ars Technica
0 Comments