Daily news

Chips in Space

by Ronald van der Breggen

Buying a new battery for your car will set you back around $100. If you need to buy one for a small airplane, you're easily looking at $500 for the same sort of battery. Why? If a plane falls from the sky the impact is much bigger than when a car breaks down by the side of the road. When the stakes are higher, the risk needs to come down. As a result, the quality standards will be raised which will make the price go up.

In space the stakes are also high. Over the past decade however, costs for space programs have gone down significantly. Launch costs per kilogram are now under $2,000 rather then $10,000. Manufacturing costs of satellites have also come down, largely due to series production in support of constellations and satellite miniaturization such as cubesats. Simply put: Access to space is now affordable and our industry is going through a boom period with more satellites launched than ever before. This in turn drives the need for more options in satellite design, critical part of which are the actual chips being used onboard the spacecraft. These chips however come with severe limitations as space is a truly harsh environment. As a result, these ‘space chips’ have nowhere near the performance as those that we e.g. have in our cell phones. Despite their low performance however, they are much more expensive! So while costs may have come down for space in general, it wasn’t because of these components; they remained as archaic and expensive as ever.

Cosmic radiation

Once we leave the earth’s atmosphere, cosmic radiation – coming from outer space and the sun – takes over and turns space into a hazardous environment in which computer chips don’t last very long. Cosmic radiation can cause what´s known as Single Event Upsets (SEUs) better known as ‘bit-flips’. They can also cause Latch-Ups, basically a type of short-circuiting, destroying the chip over time. To give you an idea, in September 1991 a Space-Shuttle mission reported 161 separate bitflips in a 5-day period. Imagine 161 ‘blue screens of death’ on your laptop while you are working against a deadline! While contrary to a Latch-Up, a bit-flip may not damage the hardware, if not properly managed it will wreak havoc to your space mission all the same.

The traditional solution to this problem is to take regular computer chips and put them through a different, more rigorous production process such that they come out hardened, now able to withstand 30 to 40 times more radiation. However, radiation hardening by process or RHBP, as it is called, causes the chips to be between 5,000 to 10,000 times more expensive when compared to components of the shelf (COTS). That makes a factor of 5 for an airplane battery look like a bargain!

As much as the traditional satellite operators and manufacturers have always used these expensive components, the new generation of space manufacturers take a different approach. For their mission they often take a slightly higher quality product than regular COTS, these would be the chips used in automotive which are ‘only’ 200 times more expensive. This puts the mission on a higher risk profile relative to RHBP chips, but for shorter missions (1-2 years) that may not be a problem. Sometimes they introduce triple redundancy which basically means that three computers work in parallel, and if one computer does not agree with the other two, then that computer is ignored and rebooted. Sure… three times the mass and volume, three times the power and three times the cost, but it works. SpaceX is well known for deploying this technology on their Falcon launch vehicles.

Satellite manufacturing may have come down in pricing, but as we have seen, it isn't because of what happened on component level. Lower costs here are merely a function cheaper, more vulnerable chips in combination with redundancy, at the expense of more mass and higher power consumption.

Developments in Radiation Hardening

What are the developments in radiation hardening itself then? Is it indeed simply a choice between the very expensive, military grade RHBP components and the slightly better than COTS, automotive products in a more redundant setup? Luckily there are plenty developments in this 1.4 B$/yr industry (marketsandmarkets.com). One of the most promising ones is radiation hardening by design or RHBD. Using this approach, chips are designed with cosmic radiation in mind, which practically means that additional hardware and detection mechanisms are built in, to take immediate action whenever a radiation anomaly is detected. Despite their large investments in the expensive RHBP technology, from their websites we learn that some of the big ‘space chip’ manufacturers like e.g. Bae Systems and Texas instruments, are already investing in RHBD. Younger companies like Vorago Tech from Austin (TX) and Zero Error Systems (ZES) from Singapore are exclusively focusing on RHBD as its design methodology brings radiation hardness well beyond those of automotive parts – even in their triple redundancy setup. All that at a much lower price point.

While the use of space components is not often talked about in the boardrooms of satellite operators or satellite manufacturers, arguably it should. Making the right decisions can lower the total cost of ownership (TCO) by spending less money while extending the life of the satellite. It can also help solve space debris issues by minimizing the risk of losing control over the satellite. And finally it can also increase the satellite’s capacity by using the extra power and space that’s now available by dropping triple redundancy. All this will help an operator to run a more successful space business, made possible by simply selecting better chips.

If I read that e.g. ZES’ technology allows for the use of COTS in space, without the risk of latch-ups damaging the chips, then satellite manufacturers and operators should really pay attention! Chips being ‘fried’ in space because of radiation, now a thing of the past? If this is true, then we're at the beginning of a revolution in space: one where from now on any satellite operator can develop any type of mission and can use any type of chip. It would truly be a game-changer.

The next revolution

Our previous revolution in space was grand in nature when SpaceX started using reusable rockets and in doing so massively slicing launch costs while putting up a great show every time a Falcon IX was launched. 

It might well be that the next revolution is virtually invisible: An opportunity to use any of the latest chipsets in space. Imagine what the International Space Station – currently running on hardened Intel 386 technology – could do if it were able to use the latest generation Intel Core i10 processors.

Now imagine what our satellites could do with that type of processing power.

It very much seems this future is knocking at our door this very moment.

Meet the author - Chips in Space | LinkedIn