Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Google AI Cellphones

Gemini Nano Won't Come To Pixel 8 Due To Hardware Limitations (mobilesyrup.com) 7

An anonymous reader quotes a report from MobileSyrup: Google's new smart assistant, Gemini, is available on multiple devices but Gemini Nano, the multimodal large language model, isn't coming to all Pixel smartphones. Gemini Nano is only available on the Google Pixel 8 Pro and the Samsung Galaxy S24 series; however, we've recently learned that it's not making its way to the base Pixel 8, according to Terence Zhang, an engineer at Google and reporter by Mishaal Rahman.

Zhang told everyone that Gemini Nano isn't coming to the Pixel 8 because of hardware limitations, but it's unclear what the hardware limitations are. Many would assume it's due to the Pixel 8 housing only 8GB of RAM compared to the Pixel 8 Pro's 12GB. That said, the Galaxy S24 series starts at 8GB of RAM and can use Nano. This must mean that some other hardware limitations are holding back Gemini Nano. Hopefully, more information will come in the future, but right now, it seems like only high-end devices will get the Gemini Nano experience.

This discussion has been archived. No new comments can be posted.

Gemini Nano Won't Come To Pixel 8 Due To Hardware Limitations

Comments Filter:
  • I consider this a selling feature.

    • I take your point, language models on embedded devices, I am sceptical on battery life, privacy concerns aside.

      To me it's a signal to skip Android 15 series phones. Wait a year and the budget Pixel will have this as a standard feature.

  • New features won't be available to the Google flagship phone launched 6 months ago. This got to be a new record of how short a flagship phone stays "flagship".

  • it's the hardware limitation of "we think this provides more value than you paid"

  • it's unclear what the hardware limitations are.

    The hardware limitations I would think would be the power of whatever machine learning processors are in the phones - just like for games you really need a GPU, for running modern machine learning models you need a device that has hardware acceleration support.

    However there is still a mystery, as according to specs I can find both Pixel 8 and Pixel 8 Pro have a "Google Tensor G3" chip. Maybe on the 8 that has more limited access to memory, but it does seem odd

The most exciting phrase to hear in science, the one that heralds new discoveries, is not "Eureka!" (I found it!) but "That's funny ..." -- Isaac Asimov

Working...