Urgent.News

What's breaking now, across thousands of outlets.

AI

Google Is Taking a Major Step Forward Toward Space-Based AI Data Centers

Google will be the first big tech company to put an experimental satellite in orbit to answer AI queries.

Google is making significant progress in its pursuit of space-based AI data centers, as revealed in a recent announcement. The primary concern with traditional data centers is their substantial power consumption, which could strain the power grid. In fact, data centers are projected to consume up to 17% of the U.S. electricity by 2030, according to the Electric Power Research Institute.

Amidst this, companies like Blue Origin and Space Exploration Technologies have proposed the concept of orbital data centers. These would utilize solar energy, minimizing reliance on Earth's resources. However, it's Google (NASDAQ:GOOG) that has taken a concrete step in this direction.

On October 1st, Google plans to launch an experimental satellite designed with Google's tensor processing units (TPUs) and solar panels. This initiative, part of Project Suncatcher, aims to eventually establish data centers in orbit to process artificial intelligence programs. The satellite, launched via a SpaceX Falcon 9 rocket, will possess the computational power equivalent to a single data center server and will operate using 1 kilowatt of solar power to process basic AI inquiries.

The satellite is designed for a year-long operation.

Written by urgent.news from Motley Fool's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Also reported by 1 other outlet

Read the original at fool.com →

More in AI

PrismML brings its tiny LLMs to Qualcomm-powered smart glasses

Prism's larger goal is open-weight AI that runs on devices and makes better use of the computing power they already have.

  • PrismML develops 1-bit Bonsai LLM for smart glasses
  • Model reduces larger AI models by 4x while maintaining performance
  • 2-billion-parameter vision and language model for real-time object recognition

More from Thursday 24 September →