WASHINGTON, Jan 29 (Reuters) - A proposed merger between Elon Musk's SpaceX and xAI, reported exclusively by Reuters on Thursday, could give fresh momentum to Musk's plan to launch satellite data centers into orbit as he battles for supremacy in the rapidly escalating AI race against tech giants like Alphabet's Google, Meta and OpenAI.

Here is what we know about space-based AI computing:

WHAT ARE SPACE-BASED AI DATA CENTERS?

Space-based data centers - still an early-stage ‌concept - would likely rely on hundreds of solar-powered satellites networked in orbit to handle the enormous computing demands of AI systems like xAI's Grok or OpenAI's ChatGPT, at a time ​when energy-hungry Earth-based facilities are becoming increasingly costly to run. Advocates say operating above the atmosphere offers nearly constant solar ‍power and eliminates the cooling burdens that dominate ground-based data-center costs, potentially making AI processing far ⁠more efficient.

But engineers and space ⁠specialists caution that commercial viability remains years away, citing major risks from space debris, defending hardware against cosmic radiation, limited options for in-person maintenance, and launch costs. Deutsche ‌Bank expects the first small-scale orbital data-center deployments in 2027-28 to ​test both the technology and the economics, with wider constellations - potentially scaling into the hundreds or thousands- emerging only in the 2030s if those early missions work.

WHY DOES MUSK WANT TO DO THIS?

SpaceX is the most ⁠successful rocket-maker in history and has successfully launched thousands of ‍satellites into orbit ​as part of its Starlink internet service. If space-based AI computing is the future, SpaceX is the most ideally placed to operate AI-ready satellite clusters or facilitate the setting up of on-orbit computing.

"It's a no-brainer building solar-power data centers ‍in space ... the lowest-cost place to put AI will be space, and that will be true within two years, three at the latest," Musk said at the World Economic Forum in Davos earlier this month.

SpaceX is considering an initial public offering this year that could value the rocket and satellite company at over $1 trillion, Reuters has reported. Part of the proceeds would go to funding the development of AI data center satellites, sources say.

WHAT ARE MUSK'S COMPETITORS DOING?

Jeff Bezos' Blue Origin has been working on technology ​for AI data ‍centers in space, building on the Amazon founder's prediction that "giant gigawatt data centers" in orbit could beat the cost of their Earth-bound peers within 10 to 20 years by tapping uninterrupted solar power and radiating heat directly into ​space.

Nvidia-backed Starcloud has already offered a glimpse of that future: its Starcloud-1 satellite, launched on a Falcon 9 last month, carries an Nvidia H100 - the most powerful AI chip ever placed in orbit - and is training and running Google's open-source Gemma model as a proof of concept. The company ultimately envisions a modular "hypercluster" of satellites providing about five gigawatts of computing power, comparable to several hyperscale data centers combined.

Google is pushing the space-based data center idea with Project Suncatcher, a research effort to network solar-powered satellites equipped with its Tensor Processing Units into an orbital AI cloud. ​The company plans an initial prototype launch with partner Planet Labs around 2027.

China also plans to create a "Space Cloud" by launching space-based artificial intelligence data centers over the next five years, state media reported on Thursday. China's main space contractor, China Aerospace Science and Technology Corporation, vowed to "construct gigawatt-class space digital-intelligence infrastructure," according to a ‍five-year development plan.