The technology giant Google revealed that running artificial intelligence models on smartphones consumes a lot of random memory.
Google was planning to provide the Gemini Nano model in the Pixel 8 Pro phone only, while the regular Pixel 8 version will be ignored due to restrictions related to hardware specifications.
Gemini Nano model
Later, Google decided to provide the Gemini Nano model in the Pixel 8 as an option for developers only, which means that users will only be able to turn it on through the developer options menu hidden in the settings, while it will be provided by default in the user interface of the Pixel 8 Pro.
In this regard, Google’s Vice President of Devices and Services Programs, Siang Zhao, confirmed that Google wants to avoid reducing the experience on the Pixel 8 phone, which has less random memory, compared to the other version.
He added that the company wants some of its AI-supported features to be located in random memory, which means that the model will permanently occupy part of the random memory to run those features at any time.
Facilitating daily plans using artificial intelligence
Google has recently added a package of updates and new features. To facilitate daily plans by people using artificial intelligence, including the Generative Search Experience (SGE) that allows users to build travel itineraries and trip ideas.
At the same time, some leaks on American websites indicate that Apple is working to increase the random memory and storage capacity in the basic versions of the upcoming iPhone 16 phones, in preparation for including artificial intelligence models capable of working in those devices.
Gemini chatbot
The Gemini robot comes with various subscription options, where users can test the company’s top AI-powered model, the Ultra 1.0, for a cost of $19.99 per month.
“Gemini” supports 3 sizes: Nano, Pro, and Ultra. In addition, it is more advanced than any other systems that support artificial intelligence. The robot also offers free options for experimentation in Gemini and Google AI Studio, and also supports interaction and the possibility of using it with various “Google” services and products.
Google recently upgraded its virtual chatbot, Bard, with a version of the Gemini model, with plans to add Gemini to widely used products including the Google search engine and the Chrome web browser, which are used by billions of people. around the world.
The company said that it trained the Gemini robot using a new generation of powerful cloud processors that can collectively train large artificial intelligence models approximately three times faster than the previous version.
For more: The most important features of the artificial intelligence assistant “Gemini” from Google
You can follow the latest news via Sayidaty’s account on Twitter
