Computing on the edge of the final frontier
On July 20, 1969, Kirk Bresniker had a one-of-a-kind front-row seat for the moon landing that other kids could only dream of.
As images of the Lunar Module Eagle navigating the Sea of Tranquility flickered on his family's black-and-white TV, Bresniker's father enthralled him with play-by-play commentary that only a rocket engine engineer would know.
Fast-forward five decades and Bresniker, a groundbreaking computer hardware pioneer and Hewlett Packard Labs' chief architect, once again has a front-row seat for a historic space mission: On Feb. 20, NASA will launch HPE Spaceborne Computer-2 to the International Space Station (ISS).
Spaceborne Computer-2 aims to build on the success of the Spaceborne Computer-1 mission, which launched the first off-the-shelf supercomputer to the ISS in August 2017. Spaceborne Computer-1 spent 615 days on the space station, running almost flawlessly and proving that the computer system could withstand the harsh conditions of space.
Scaling up for its return voyage, Spaceborne Computer-2 will bring enhanced edge computing and AI capabilities to the ISS, boasting twice the compute power of Spaceborne Computer-1 and reducing the time to insights from months to minutes.
I spoke to Bresniker about growing up in a "space family" and his insights into how Spaceborne Computer-2 could help usher in the next great chapter of space exploration.
Tell me about your father's involvement in the Space Race?
My dad graduated Santa Clara University in 1959, right at the height of the buildup, and went into aerospace engineering at United Technologies, where he worked on the solid rocket booster systems.
We were, and have continued to be, a space family. He was always crisscrossing the country every couple of weeks to go to Florida or down south to Vandenberg Air Force Base to watch one of his "birds" fly. One day, he brought home one of the first pictures of Mars from the Viking lander, and he put it up on the wall because it was his engines that helped accomplish that mission.
Were you interested in what he was doing as a kid?
I was. When I was little, I would watch my dad working on getting his master's degree—class before work, homework after dinner, using his handbook of mathematic tables and his slide rule (a wedding present from my mother). I just remember staring over his shoulder at the table of integrals and those beautiful notations. Of course, I was only a couple of years old, but still, there was something about what he was doing that was fascinating to me. I still have that book at my desk.
My parents nurtured my interest. When I was a little bit older, my dad got a gigantic surplus calculating machine that they were getting rid of at work. It was the size of a washing machine. He brought it home and said, 'There's the toolbox. Tear into it, take it apart, and make something out of it.' At every opportunity, he was inviting me in. 'What do you want to try? What can I do to set you up so you can tackle this stuff yourself?'
I loved it. There wasn't anything that my dad didn't tackle. He'd look at a problem, write it out, and work on the solutions. He was tinkering, trying things out. I think he certainly passed that on to me.
What's exciting to you about the Spaceborne missions?
The ability to get into space in a way which is sustainable. So far, we've gathered a bunch of resources and tried to lift them up into space, to escape the gravity well. This is about, can we actually capture energy in space, take materials from space, manufacture in space? That means we're moving toward the ability to sustainably fund our aspirations in space. And Spaceborne grows that incredibly.
When I think of space, I think of the fact that basically we live from sea level up to a couple thousand feet that's just a very thin envelope around the planet. All the things we depend on either come from one of two places: from the sun for energy, or the earth for the raw materials. And that's all we have. When you think of that thin shell of life that we have, and whether by the foolishness of our own hand or the fist of God, there's every chance that that fragile ecosystem could be destroyed.
I want to see us succeed in space because I want to make sure that mankind has a refuge other than just what we can have on the surface of the earth. And I want to see Spaceborne succeed, for the same reason I do the work I do to solve problems here on earth: I want to see that same capability to gain insight to our most challenging problems from breakthrough information technology brought to space
Those are lofty goals. How will Spaceborne help us get there?
There are challenges in space. On earth, the thin atmosphere does a lot of good for us. The earth's magnetic field seals us from the solar wind, the high-energy particles that flood space, including the low Earth orbit of the ISS. When systems designed for Earth are subject to those harsher conditions, they are likely to fail, either instantaneously because you get hit with a cosmic ray or over time, because the harsh environment of space begins to degrade materials that on earth will last forever.
The technology we have designed and built within the relative comfort of our earth, can that cutting-edge technology survive and function in the harsh environment of space? That's really what we're learning to do with Spaceborne: to adapt today's most efficient, compelling technologies to understand how we can make it work up there and work reliably enough that we can give it our toughest or most important problems.
This isn't just about today's mission in low Earth orbit. It's about tomorrow's mission to return to the moon and the day after tomorrow's mission to explore Mars and hopefully on to the asteroid belt.
What challenges to those ambitious future missions will Spaceborne Computer-2 help us solve?
The pipe is relatively small to get data back from the ISS to Earth, let alone the moon or Mars. I want to have the same kind of rich, data-driven machine learning, inference-capable, trained insights that we have on earth—that will allow us to do things like power autonomous vehicles—happen in space. And we cannot rely on a data center on earth to power that because the round-trip time is too long.
Spaceborne is letting us gather this incredible volume of data and center it in space. We have to move the compute out to where the data is. That means it has to survive and thrive in that harsh environment.
A lot of naysayers didn't think Spaceborne Computer-1 was going to last a week, let alone 651 days. Were you surprised at how successful the first mission was?
I was very happy to see the results. You run an experiment like this because you believe there's an opportunity, so any answer is rewarding. But it's definitely thrilling to know you're on to something. The team was able to apply clever software and error-correction techniques on top of today's state-of-the-art hardware and came up with a system that was able to not only survive but thrive to do its job day in and day out.
That's really what I was so excited about for this team. And I'm very excited to see them continue on. Having proved it once, what more can you do? Can you make the computer even smaller, even lighter, use even less power? What can you do next? How can you go even further?
That's really what we're learning to do with Spaceborne: to adapt today's most efficient, compelling technologies to understand how we can make it work up there and work reliably enough that we can give it our toughest or most important problems.
This article/content was written by the individual writer identified and does not necessarily reflect the view of Hewlett Packard Enterprise Company.