Home

Orbital datacenters are a pie-in-the-sky idea: Gartner

Analyst firm Gartner thinks talk of placing datacenters in space has reached “peak insanity,” because orbiting facilities can’t be run economically or satisfy demand for compute power on Earth.

“Datacenters in space won’t analyze data on Earth for Earth applications for decades, if ever,” states a report published this week titled “Orbital Datacenters Won’t Serve Terrestrial Needs, So Focus on Earth,” penned by distinguished VP analyst Bill Ray.

“Companies are wasting money by pouring funds into the orbital data center ‘bubble’ because the economics do not work,” the analyst wrote. “This is due to the prohibitive costs of launching hardware and the immense technical challenges of cooling these orbital datacenters in the vacuum that is space.”

Ray noted that orbital datacenters must be able to survive “extraordinary swings in temperature, from 100 degrees kelvin to 400 degrees kelvin” and that doing so will require “specialized components (like solar panels) that cost roughly 1,000 times more than their terrestrial counterparts.”

The analyst also thinks orbiting datacenters will need specialized cooling equipment – probably ammonia piping that shifts heat from computers to radiators as deployed on the International Space Station.

“Maintaining such complex infrastructure would require a fundamental shift from sending astronauts into space to sending engineers to perform necessary maintenance, a capability that does not currently exist,” Ray wrote.

He also thinks that proponents of using lasers to transmit data from orbiting compute clouds to Earth have it wrong, because actual clouds disrupt transmissions and make for inconsistent data transfer speeds.

Ray doesn’t entirely dismiss the concept of orbiting datacenters.

“The technology will ultimately evolve to support only data produced in space for consumption in space, such as processing satellite imagery to remove cloud cover before transmission or managing complex mesh routing for communications,” he predicted. And he thinks that research conducted to make space datacenters work could have benefits for terrestrial bit barns.

But for now, he believes those seeking additional datacenter capacity should focus their efforts on Earth.

“Product leaders face the very real possibility of underbuilt terrestrial data center capacity if this ‘bubble’ of space data center hype lasts for several years,” he wrote, and suggested instead considering the underwater datacenter technology pioneered by Microsoft, or building bit barns in the Arctic, Iceland or the deserts of Saudi Arabia. ®

Source: The register

Previous

Next