Advanced Computing in the Age of AI | Friday, March 29, 2024

Microsoft, HPE Bringing AI, Edge, Cloud to Earth Orbit in Preparation for Mars Missions 

The International Space Station will soon get a delivery of powerful AI, edge and cloud computing tools from HPE and Microsoft Azure to expand technology experiments aimed at preparing NASA for launching future crewed exploratory missions to Mars.

The new equipment and software, including HPE’s specialized, second-generation Spaceborne Computer-2 (SBC-2), will mark the first time that broad AI and edge computing capabilities will be available to researchers on the space station, Tom Keane, Microsoft’s vice president of Azure Global, wrote in a Feb. 11 post on the Azure blog.

The new hardware, software and services are scheduled for launch to the ISS at 12:36 p.m. on Feb. 20 aboard Northrop Grumman’s 15th (NG-15) Commercial Resupply Services cargo mission. The NG-15 mission’s launch from the Wallops Flight Facility at Wallops Island, Virginia, is contracted by NASA to bring needed supplies.

The new SBC-2 computer that’s heading to the space station follows the original Spaceborne Computer-1 that was sent to the ISS in 2017 as part of a validation study to test it in the rigors of space aboard the orbiting laboratory. SBC-1 returned to earth in 2019 after completing its mission. Both Spaceborne Computer-1 and Spaceborne Computer-2 are sponsored by the ISS National Lab.

Dr. Mark Fernandez of HPE

SBC-2 will bring ISS researchers a wide range of new capabilities they did not have with the original Spaceborne machine from 2017 to 2019, Dr. Mark Fernandez, solution architect for converged edge systems at HPE and principal investigator for SBC-2 told EnterpriseAI. Technological advancements in AI, cloud and more will provide more possibilities in the new machine for ISS researchers, he said.

“Hardware-wise, we're sending up the HPE Edgeline Converged EL4000 Edge system, which is purpose-engineered and built to operate on the edge and take advantage of AI and ML capabilities with its onboard Nvidia T4 GPUs,” said Fernandez. “These are enterprise-class, commercial off-the-shelf servers that go into data centers.”

Featuring CPUs and GPUs

The Edgeline EL4000 servers will use Nvidia T4 GPUs for AI and machine learning, image processing, video processing and other tasks. Previously, the first SBC-1 used CPUs for those tasks. The latest SBC-2 will include CPUs and GPUs to allow for comparison performance experimentation in space.

The 1U boxes insert into standard data center 19-inch racks on the ISS. The racks are then inserted into lockers aboard the ISS to hold them securely. Also provided are an enterprise-class compute node, HPE’s ProLiant DL360, for intense compute requirements, said Fernandez.

HPE's Spaceborne Compuer-2

For the second generation of the SBC, NASA asked HPE to send up twice the compute power of the original version, said Fernandez. “So, we're sending up twice the number of servers. You'll see two lockers and each contains two servers.”

“One is a CPU-based Intel server for those that love Intel and traditional computing, and we'll have a GPU-based Edgeline server for those that are doing image processing, artificial intelligence, machine learning, etc.,” said Fernandez.

NASA asked for double the computing power in the SBC-2 so the agency can continue its work toward sending humans to Mars, he said. SBC-1 was a proof-of-concept device for 18 months and now the new SBC-2 will be tested to see how it reacts to two to three years in space to accommodate a mission to Mars, he added.

Azure in Space

The Azure cloud capabilities will be used with the machines to allow experiments with getting data back and forth from the ISS to Earth as quickly and efficiently as possible, said Fernandez. Such data transfers are done today using existing NASA technologies.

“The ISS is only 220 miles up in Earth orbit, but the networking is circa-1980,” said Fernandez. “We have speeds of two megabits a second up and down to the ISS. I have 50 megabits a second in my home.”

Increasing those speeds will be critical for Mars missions, he said.

“Microsoft is enabling that, and they have aspirational plans to come up with some AI and machine learning that we'll look at,” said Fernandez. One idea they will look at is running data on SBC-2 and then sending small amounts of data back to Earth, and then comparing that to bursting data to Azure and seeing what works faster.

“We're sitting right on top of the same NASA network, but we're going to encode and compress messages back and forth in order to take the most advantage of that two megabytes per second,” he said. “I have a brilliant scientist who is going to run the same experiment on CPUs, on GPUs and in the cloud. And he will report back to the community, if you have this type of data, it's best if you process it this way because we are given those three options.”

The experiments will begin after the equipment arrives at the ISS and following their installation and setup. Those tasks are expected to take some time to complete, including several days for the cargo mission to arrive at the space station. “We've got three pre-canned experiments for three different users that we're going to hope to fire off right away,” said Fernandez.

How Azure Views Its Mission Aboard the ISS

“The crux of this work is about making the capabilities of Azure available to astronauts, space explorers and researchers to learn and advance science and the use of the cloud to support their goals,” a Microsoft spokesperson told EnterpriseAI. “Through this project we will be able to continue to gain knowledge on how we can best support the science and research community, wherever they are, on and off the planet.”

With SBC-2, Microsoft’s research and Azure space engineering teams are “evaluating the potential of HPE’s [space-based] state-of-the-art processing in conjunction with hyperscale Azure, alongside the development of advanced artificial intelligence (AI) and machine learning models to support new insights and research advancements,” the spokesperson said.

That includes weather-modeling of dust storms to enable future modeling for Mars missions, plant and hydroponics analysis to support food growth and life sciences in space, and medical imaging experiments using an ultrasound on the ISS to support astronaut healthcare. Also being created is a platform for the development and testing of hybrid edge-cloud environments before contributing additional experiments to the ISS.

“We are exploring the potential of empowering exciting new experiments that leverage the far-reaching potential of the cloud in conjunction with the HPE Edge capabilities,” the spokesperson said. “To date, researchers have had to often limit the scope of their study to what computational resources they had available to conduct their research.”

Using bursting capabilities with Azure will add to future capabilities, according to Microsoft. “Bursting down to the cloud provides access to more computation/resources than can be hosted in the ISS, while leveraging SBC-2’s power and proximity at the edge,” the spokesperson said. “We are excited to empower others, even in space, to be able to leverage the power of Microsoft Azure - making it possible for astronauts, space explorers, and researchers to think big as they tackle their toughest questions.”

 

EnterpriseAI