I like this sort of sponsorship; "Hey, we did a cool thing, let us pay you to come out and see it all and share how cool it is, to get our brand out there."
Couldn't agree more. I mean, it's a flex from Kioxia for sure, but damn, it's a big flex. I'd love to be able to migrate to Kioxia SSD's for my NAS, even if they're monstrous overkill.
Yeah, some years ago If you said Kioxia I would ask which disease is that, and if I wanted a good SSD I would go Samsung or Intel, now I know about them.
11:41 One correction. The cooling loops exchange heat with the External Thermal Control System radiators (the two sets of 3 radiators closest to the modules along the truss) The Photovoltaic Thermal Control System Radiators are the 4 radiators mounted between each pair of solar arrays. They only provide cooling to the electronics for their respective solar arrays. This makes sense because they are mounted outside of the 2 Solar Alpha Rotary Joints which spin the 4 port and starboard solar arrays 360 degrees every orbit. It’s hard to pass a fluid through a rotating joint.
@@discorddiscord2647 i think he actually mentioned this on wan show; he's in a replica of the iss at one of hp's offices. or smth like that 17:11 oh there you go
OMG ! they put a "do not stir" label at 16:01 , above the "o2 cryogenic tanks" label. For those not as geeky as me, its a reference to the Apollo 13 lunar mission, that suffered an explosion due to a short circuit...while stirring the O2 tanks ! (Fortunately no one died, its a really cool story, there's also a movie about that)
So, in the once upon a time my grandfather worked for JPL, and I did in fact get to see a couple of the ISS modules at Cape Canaveral before they launched. And I can tell you... they basically looked like big white rectangles. Researching the tech behind them is honestly more fun. But as far as computers surviving in space... my grandfather did work on two such computers that are still working today, more than 40 years after they launched... the Voyager probes.
That is so cool. Space is so incredible and the folks like your grandfather than helped us learn so much about it are all legends. They deserve way more credit and appreciation from us than they get.
but those computers were no were near as powerful, and were specially designed for radiation, these were COTS servers with no radiation protection of any kind.
there are a few things called satellites up there also. not surprising they also have computers up there. the big problem is the ISS seemed to be planned and executed with long slow and dangerously bad design and practice. everything from the station not being completed till after it was supposed to have reached end of life. they run at low pressure resulting in converting super fit & healthy humans into very weak people. they could dissipate heat from computers to heat the station.
@@chrisoakey9841 Utter nonsense. "they run at low pressure resulting in converting super fit & healthy humans into very weak people." That in no way weakens people, it's lack of gravity that does it, same as bed-ridden hospital patients. "they could dissipate heat from computers to heat the station." The problem is Not heating, it's cooling. Without sufficient cooling, the ISS would overheat and cook teh astronauts alive from getting too hot. The computers also need cooling, and so are limited by the total ISS cooling budget, which is limited by the size and number of their radiators. Since the ISS continued longer than planned, they kept bolting on new unplanned additions. things take years to design, test, validate,etc. before flown, and so something planned in the 1980s and initially designed in the 1990s is going to be obsolete 35yrs later, who knew. The ISS was initially developed in the days before personal computers, internet, and other technology being commonplace. Even most electronic calculators in teh early 1990s were still very crude devices, with 2-line calculators and more scientific computing features coming along later.
And/Or, Because it took so much time to build things and most of the NASA space tech was designed and engineered decades ago, They can't just update/replace to accumulate new tech.. Like for the ISS.
@@mozzjones6943 and yet they can't even follow their own guidelines for making a rocket go figure. joking aside while the getting things to space and having it work issue is a factor most of it's Bureaucracy.
well government has never been good at moving fast. Private industry should be in charge of the tech and integration. I mean a us defense contractror charges 90k dollars for a very small bag of bushings. they charge too much because the government pays with tax dollars.
14:47 for anybody feeling nostalgic, remember AOL and dial-up. You would have to wait 20 minutes for a single song to download. Now you can relive the experience, all you have to do is become an astronaut.
How about sending a >1MB file through MSN? That would also take minutes at best and had a very high potential to fail, at which point you had to start all over again. Sounds very much like what they're dealing with in space today :D
Wow! This has got to be one of the coolest videos I’ve ever seen on UA-cam! Great to know how the ISS operates and its constraints. It’s nothing like I expected! Please do more videos on space tech if you can 😊
I work in the space industry building rocket engines. Super awesome to see you pump out content that I can relate too on a personal level. Thank you for making this video and I am excitedly waiting now for more space tech content :)
Not really surprising. Think about a submarine - the other kind of pressure vessel we use as a species that deals with dramatic changes. The ISS deals with a total pressure differential of ~1 atmosphere. Relatively speaking, it's easy to keep the thing as one solid piece. The difficult part is the lack of atmosphere. Take a hinge - like a standard door hinge, two pieces of brass or whatever, that slide together. In space you would find it has the awful habbit of sticking after you open it a few times, and by sticking I mean suddenly welding itself together as the oxide layer is rubbed off, exposing bare smooth metal surfaces that don't know any difference than to presume they are a single continuous piece of metal and suddenly: You can't easily move the hinge. This isn't a problem on earth - any oxide layer that does get removed, quickly reforms. But where you have no atmosphere? The problems in Space aren't so much about bleeding edge technology - it's about solving problems that you didn't know would be problems. This is why Reliability, trumps basically all else. You need to know it will work.
It's like seeing explosion proof secruity cameras in a companies product line, then thinking they must be some crazy kind of armored camera. Then finding out it just means they won't cause an explosion not that they would survive one.
They launched a hard drive with a super capacitor, but accidentally launched an experiment that tested what happens to unshielded super capacitors when bombarded with gamma radiation. The ISS will ALWAYS manage to gather data about something.
@@noobulon4334 The results were simple: SSDs in space fail much faster than on the earth, as the super capacitators break faster and therefore the SSD gets overcharged during saving procedures earlier compared to earth. Solution: regularly backup your data - that's why they have this big server with longer living capacity, so that they can save the data from the experiments all the time and then send it to earth in big packages.
True science is just as excited about a bad result as a good result. Because it's still a result that they can learn from. And let us never forget just how many amazing discoveries were made by accident while trying to learn something entirely different. Accidental results are some of the best we've made.
Thank you so much for the video!! I think this may be my personal favorite. So cool to see the practical applications in such a unique and challenging environment.
You just know he used this opportunity to put a few into the chain there. Hoping that one of them might actually make it up to space and get on one of the live streams they do.
This is so awesome. While watching this, my eyes were sparkling so much, one might have seen the light from the ISS. I love this video. Very well-done.
Hi I'm working in aerospace, I can talk about just a couple of things: 1. People think of space as cold, but it might not be as cold as you think. Temperature in low Earth orbit (LEO) can be -60 to +120 C depending on how much direct sunlight you're receiving. Linus' point about having no air to convect heat away is a pretty big deal - that's why you need massive radiator panels. 2. NASA and incumbent space agencies tend to be risk-averse and want to see SIGNIFICANT verification of functionality at every level, particularly when humans are involved. That's probably one reason why they don't use Starlink - the technology is simply too new and maybe there's some technical or bureaucratic inertia to overcome. It's also why tech in space tends to be pretty old, even if it's launched today. US agencies like to see flight heritage. So even (as a silly example) if Ryzen 7600 is better than Intel Core 2, if these agencies see that Intel Core 2 provides everything they need, they're more likely to pick the Core 2 over Ryzen 7600 because they KNOW it'll work. This is compounded by the fact that the rad-hardening process takes a ton of money and time, so you often end up sending surprisingly old tech into space. Newer commercial companies and smallsats are trying to use newer tech though. 3. Power budget is super important. One reason that it's so tight is because solar panels degrade over time, especially in space, so the power available to the ISS is decreasing over time. 4. Bit flips are MUCH more common than on Earth, that's why rad-hardening is important. I had a professor whose research specifically tried to ameliorate the effects of bit-flipping. 5. Not necessarily tech related but it's true that space vehicles need to balance physical practicality versus human preference, e.g. convenient orientations versus "needed" orientations like Linus mentioned.
@@Deinorius They did use Starlink to link to the recent Starship test flight, but that was lower than the satellites. I guess at some point SpaceX will put a dish in/on a Dragon capsule to test it further up.
i work in a company that launches cubesats and we just had a radiation incident a week or so ago! can bus traffic to the power management board completely dead and watchdogs not triggered, radiation’s bad for the little dudes lol
@@Deinorius they have some sort of communication capability in space. I'm not sure off the top of my head what it is - but we know they're expected to be able to respond in case of collisions. Whether that can be repurposed for communication with Earth, I dunno
Why don't they use Radioisotope Thermoelectric Generators instead or in combination with solar? They could reliably generate hundreds of watts for decades.
This brings back memories when i was a test engineer at a satellite company. Nearly every circuit had a duplicate on the same chassis. Then they would put two of those in which is essentially 4 copies. They have to guarantee a satellite to work for 10-15 years without being able to repair it.
That scene from MSB of Arnold taking his helmet off on Pluto, lives rent free in my head. one of these few episodes I ever actually saw (was probably with school tbh)
A friend of mine is a physicist who used to work at CERN and, as part of her job, assembled sensors for the particle collider. And despite the sensor's specific purpose being to measure the products of a collision, one of the issues they had to deal with was these particles causing bit flips and other issues when passing through the chips inside the sensors. So what she told me they did was to irradiate the chips before putting them in because this somehow increased protection from the effects of the particles passing through. Or better put less susceptible to damage. She tried to explain it to me what happens in detail but this was long ago and I didn't understand half of it, so I might be talking BS here. Though I wonder if NASA/ESA also do something similar to their chips. But I guess this would only be useful for heavier particles than gamma radiation, which I guess is the main problem in space. All that being said, computing in hostile environments is both very interesting and potentially very frustrating as entropy slaps you in the face.
Physicist here, you pre eradiate them to get those that are particularly susceptible to die before you put it into the heart of a machine you can only service every few years. The procedure does not make anything more resilient, just the surviving units are less likely to fail later on.
@@XanatosDavid Ah okay, that makes much more sense. Then I remembered it wrong. I just remembered something about irradiating chips and some presentation slides she sent me years ago. Though I looked through CERN's DB and found them, and you are completely right, it's just part of a testing campaign. In case you're interested (they mention "irradiation qualification of ASICs"): indico.cern.ch/event/774201/contributions/3429235/attachments/1875763/3088566/iWoRiDtalkDette2019.pdf Though since you're here and could provide some background knowledge: Since this is basically just testing/sorting chips, wouldn't this also make sense for any kind of ASIC/CPU/GPU NASA sends up there? Or do they do that already?
@@XanatosDavid Somehow my last response was deleted, probably because I linked to a PDF from the CERN document server with slides my friend sent me back when she told me about this. But thank you for explaining and of course you are completely right and I remembered it wrong, it was basically just done to sort out the chips that would be most affected by radiation. In the slides she sent me back then she called it "irradiation qualification". But since I have another physicist to ask now: Do you know if NASA performs similar tests with the computers they send up? Would make a lot of sense I guess. Oh and in case anyone wonders, the PDF can be found on the CERN document server, it has ID 2686279.
@@B20C0 One of the points of this project was to send off the shelf hardware. Most computers that have been sent up before are treated like so or custom made. The spaceborn project's whole point was "modern enterprise hardware has ECC built into every layer how would it work in space?"
Something like this is actually common practice for all industrial grade (and military and aerospace grade of course) electronics. Most component failures tend to occur early, so the parts undergo a stress test inside a climatic chamber, cycling through the temperature and humidity ranges the parts are specified for. It's called a "burn-in". This is of course expensive, but it's a lot less expensive than fixing a train stuck in some tunnel in the alps or a plane crashing.
Sometimes I see the collabs that LTT does and think "hmm. Strange. Ok though, they're cool." And then I see a collab like this and it's like wtf man where did these connections happen
Fun fact: the computers on board the Perseverance and Curiosity mars sister rovers are 30+ y/o designs chosen because they are radiation hardened, and if you think conputers on the ISS are hard, keep in mind that the farther out you go, the harder it gets
@@furinick Only that neither Apple nor NASA would ever attampt to send any Mac-Based system to space. Especially now when there's esentailly no way to swap the OS on Apple Silicon Macs. There is a lot of iPads on the ISS though. Both personal and scientific iPads to track data etc. Funny enough, it was Russia that started to ship iPads to the ISS as enterntainment devices to replace aging iPods in 2011. NASA has since adopted iPads as well though. Sometimes in Crew Dragon launches you can see iPads starpped to the Dragon Flight Suites.
They're actually HP Z Books, and they're old because certifying them for flight is an expensive endeavor, so NASA does a bulk lifetime buy of the laptops and spare parts all at once.
Hate to buzzkill, but that's based on an oft-repeated misquote. Neither of the Apollo 13 astronauts involved said that. It was closer to "Houston, we've had a...".
Speaking as someone that helped put about 150 satellites into LEO, in my opinion, space hardening the hardware really only applies when you intend on a super long lifespan - which means you also have no reliable de-orbit plan (so you fully intend on being space debris). For something with a lifespan of under 10 years, you just need some redundant systems and really good error checking in your software. Software based hardening is 100% a thing and way cheaper.
Always depends on the workload, but it is crazy how many workloads can fit in the Rancher SUSE "Cattle not pets" model for services where you build them to deploy as stateless as possible to allow for processes to spin up when needed to process data and crash with no loss if they go down. Again redundancy, plan for failures to happen, and then you can start to tighten your tolerances to save costs and not just to get it to work at all!
I think I heard somewhere that you can have multiple identical CPUs executing in parallel, that way either you can detect errors and either recalculate that, reset, or if you have 3+ CPUs then you can fix one using the others.
Or when human lives depend on it. When a satellite mailfunctions you scream, write it off and send a new one. When a person dies you have a lot more questions and issues to deal with. During the "space race" it was kind of a war with russia and some lives lost were acceptable. That's not the case anymore. That's why things are over engineered and thorougly tested. Because there is no triple a or plumber in space.
The company I work for used to make a peristaltic pump for experiments on the space shuttles. It was made to be smaller, lighter, and more energy efficient than any other pump we made at the time and had to be dumb enough not to rely on anything that radiation could kill.
LEO is pretty easy, below the Van Allen belts. Even laptops work there. GEO is much more fun. Pretty rare to find a processor that works there. I had to design my own but that was 30 years ago.
The U.S. government has outsourced everything to private contractors who use research from public colleges and universities, and actual NASA is looking like some kind of retro future aesthetic using off the shelf server parts.
the private sector computers are far more advanced than anything the gov has developed. Making modern computers is not easy, and the US gov doesn't have the people, nor the equipment, nor the innovation to do it themselves. And if they did, without having volumes to sell, it would be ridiculously expensive for a single computer.
NASA has *always* outsourced a ton of stuff to private contractors. e.g. take the Apollo Guidance Computer (AGC). The MIT Instrumentation Laboratory designed the computer's hardware, OS and UI. Raytheon did the manufacturing. Fairchild Semiconductor pioneered the first mass-producible ICs in 1960, and the AGC was the first computer to use them. Of course these were still primitive-one flatpack chip roughly 20mm² might contain two 3-input NOR gates, for a total of 6 transistors. The AGC's memory was even wilder. Its RAM was magnetic core memory (typical for the time) but its ROM was core rope memory. The AGC had 6 core rope modules. Each contained 512 iron/nickel cores (donuts 0.249" in diameter, about 6.3mm) and 192 "sense" wires. For every core, each sense wire either threaded through it (representing 1) or around it (representing 0). Cores had additional wires for address selection and reading out. All of this threading was done by hand-Raytheon had a factory full of needleworkers threading data into the ROM, bit by bit. They did this because even though core rope memory was incredibly labour-intensive to make, it had unparalleled information density, about 5× that of the magnetic core memory used as RAM. When going into space, mass and volume are paramount. It also happens to be resilient to bit-flips ;)
You're too young to remember, but the first laptop in space was by Grid Systems. Their product, the Grid Compass 1101, was introduced in 1982 and became notable for its rugged, clamshell design and magnesium case. This model was used by NASA on the Space Shuttle missions, starting with STS-13 in 1983. They were VERY expensive as I recall. The computer featured an Intel 8086 processor, a 320 × 240-pixel electroluminescent display, 340-kilobyte magnetic bubble memory, and a 1200 bit/s modem. Devices such as hard drives and floppy drives could be connected via the IEEE-488 I/O (also known as GPIB or General Purpose Interface Bus). This port made it possible to connect multiple devices to the addressable device bus. It weighed 5 kg (11 lb). The power input is ~110/220 V AC, 47-66 Hz, 75 W.
Idk if you've done a video on Mars rovers or their computers (a quick UA-cam search suggests not 🤷♂️) but the radiation hardening required for those sattelites and computers that go beyond LEO are insane. For instance, the Perseverance rover launched in 2020 has the equivalent of a late-90s PowerPC Mac CPU. The ISS and other LEO sattelites don't need this because they are below the Van Allen radiation belts that capture most of the sun's harmful radiation. On another note, I don't think the wider public realizes how difficult it can be to change hardware out in space. Every spacewalk takes multiple hours and you only get half of the work done that you could here on the ground. And every upgrade (like adding Starlink antennas) would require a spacewalk. The work that goes into computers and equipment that goes into space is amazing. Awesome video, love that you could work with NASA and bring light to a wider audience the awesome things they do. 😁👍
not forgetting the possible hacking of any internal systems would be a very good reason to keep the link private, wich in the case of the iss are almost all vital systems , life support,power, orbital control...ect..
Nowadays, you can use Starlink anywhere on Earth... which doesn't help the ISS. The Polaris Dawn mission is supposed to conduct high-altitude tests of Starlink laser transcievers, which could be more useful.
Very impressed with this video. I know Linus knows his shit, but for him to talk this smoothly about the gear and the onboard systems in an engaging and fun manner yet highly informational way takes a fair bit of skill. Well done LTT on this video. Great sponsorship video. Not every video needs to home/business PC related. Now see if someone who supplies Cheyenne Mountain will sponsor you and get you access to their super computers. lol
I started using Kioxia about 6 years ago for my main hard drive brand in 6 rigs and haven't been let down by them yet for performance, reliability and price.
My ex girlfriend has a very good friend who works at NASA. He is the a main boss. He is in charge of exploratory machines. He's not the coffee guy. We became good friends through an online game she introduced me to. And we've been good friends ever since. He once told me that they were looking for a chip set that would remain stable and constant no matter how long it was left turned on for. NASA had problems creating one until SONY popped up in one of their lectures... Now most modern spacecraft use the PS1 chipset or an updated version of it. It can be left on indefinitely without overheating or failing.
My favourite space story is the US spent more than a million dollars to make a zero gravity pen for their astronauts and Russia sent their astronauts with a pencil 😂😂😂
10:59 I like the little alien floating by the window. I was wondering the Starlink question before it got conveniently answered. Though maybe it's not worth it if the ISS retirement plans go through?
Having just wrapped up my Master's thesis, which dealt with working within the constraints of edge compute systems for AI in space deployments, this was great timing and a very enjoyable video. Great work on really putting into perspective the constraints one has to work within when dealing with computer systems in space. The university I attended has a lot of projects under NASA funding, so its always nice to hear that NASA's big ISS projects have to deal with many of the same sorts of constraints that we have to put up with on our smaller space platforms.
The Morse code on the Spaceborne Computer 2 welcome page 15:03 says "GREATINGS FROM THE ISS". Also, the graphic on the left is the Silicon Graphics logo. Which has Mike, Dave, Carrie, Mark, Rob, and John hidden in it. I expect one of them can explain all of these things.
The original Spaceborne Computer was made by HPE in 2017, right after they bought SGI, and NASA, SGI, Cray and HP have a long history. I don't know the orgchart but this made me grin.
Correction at 15:20 - This is meant to say "TDRS Satellite" instead of "TRDS Satellitte."
No comments ? Lemme fix that
TDSR?
If only it was Tardis. That would be my favorite satalite.
Like the TARDIS?
Did u lot get to play with that DVD Rom Drive behind linus its got its own volume button 2:52
I like this sort of sponsorship; "Hey, we did a cool thing, let us pay you to come out and see it all and share how cool it is, to get our brand out there."
One of the best kinds of sponsorships!
I hope we see more of this kinda stuff, it's cool to see all the fancy space tech!
Couldn't agree more. I mean, it's a flex from Kioxia for sure, but damn, it's a big flex. I'd love to be able to migrate to Kioxia SSD's for my NAS, even if they're monstrous overkill.
I agree. I just hope this doesn't age poorly like Linus's trip to Intel in the terrorist state of Israel
Yeah, some years ago If you said Kioxia I would ask which disease is that, and if I wanted a good SSD I would go Samsung or Intel, now I know about them.
It’s very hard to throw a computer that high and have it still work.
Haha bro 🤣 best world pc
It’s very hard to throw a computer at all and have it still work
LEXI! Also so true.
wow a random appearing lexi
Drop a computer and have it fall that far up*
Many kerbals have died from "Computer bit flips"
"Many kerbals died to bring us this information."
@Harsh85114get a life
@@samiraperi467 A quote from Mon Kerbma
Does that explain 11:00?
(Yes, I know it Jeb. Maybe a long lost ancestor or whatever.)
That's a beautiful way of describing a rage quit right there.
As an actual engineer who deals with SWAP and the difficulties of aerospace environments, I will say this video is well made and pretty accurate.
do you have any insight on why there isn't any shielding on the computer? would you need a meter of lead or something like that?
Because effective shielding is very heavy. Especially for neutrons and Gamma rays. It takes a lot of material to do that.
male
And you been to outer space. 😂
@@evelynkieraivanova5404 how does one go in this field?
Seems interesting
11:41 One correction. The cooling loops exchange heat with the External Thermal Control System radiators (the two sets of 3 radiators closest to the modules along the truss)
The Photovoltaic Thermal Control System Radiators are the 4 radiators mounted between each pair of solar arrays. They only provide cooling to the electronics for their respective solar arrays. This makes sense because they are mounted outside of the 2 Solar Alpha Rotary Joints which spin the 4 port and starboard solar arrays 360 degrees every orbit. It’s hard to pass a fluid through a rotating joint.
I can't believe they built a fake sound stage to cover up the fact that Linus went to the space station.
I’m so glad I didn’t read this comment before watching the whole thing.
He fit into the hand luggage
They did a really good job simulating gravity, wonder how they did that.
@@gcewing linus just needed 2 very small electromagnets to stick to the surface
He actually landed on the Moon too.
Kudos to Linus for going to space without space suit just for the thumbnail
Yeah props x69,420
oh what a middle aged man would do for views right
Wait he’s actually in space or is he in a rocket on earth that will transport this equipment ?
@@discorddiscord2647 i think he actually mentioned this on wan show; he's in a replica of the iss at one of hp's offices. or smth like that
17:11 oh there you go
Not to be *that* guy, but isn't this an ISS model at NASA
OMG ! they put a "do not stir" label at 16:01 , above the "o2 cryogenic tanks" label.
For those not as geeky as me, its a reference to the Apollo 13 lunar mission, that suffered an explosion due to a short circuit...while stirring the O2 tanks ! (Fortunately no one died, its a really cool story, there's also a movie about that)
I'm glad someone else noticed that!
16:17
what movie?
@@tomikun8057 the mission was Apollo 13, the movie was also Apollo 13...
but then, maybe you're joking....
@@tomikun8057 Apollo 13
So, in the once upon a time my grandfather worked for JPL, and I did in fact get to see a couple of the ISS modules at Cape Canaveral before they launched. And I can tell you... they basically looked like big white rectangles. Researching the tech behind them is honestly more fun.
But as far as computers surviving in space... my grandfather did work on two such computers that are still working today, more than 40 years after they launched... the Voyager probes.
That is so cool. Space is so incredible and the folks like your grandfather than helped us learn so much about it are all legends. They deserve way more credit and appreciation from us than they get.
True legend!
but those computers were no were near as powerful, and were specially designed for radiation, these were COTS servers with no radiation protection of any kind.
there are a few things called satellites up there also. not surprising they also have computers up there. the big problem is the ISS seemed to be planned and executed with long slow and dangerously bad design and practice. everything from the station not being completed till after it was supposed to have reached end of life. they run at low pressure resulting in converting super fit & healthy humans into very weak people.
they could dissipate heat from computers to heat the station.
@@chrisoakey9841 Utter nonsense.
"they run at low pressure resulting in converting super fit & healthy humans into very weak people."
That in no way weakens people, it's lack of gravity that does it, same as bed-ridden hospital patients.
"they could dissipate heat from computers to heat the station."
The problem is Not heating, it's cooling. Without sufficient cooling, the ISS would overheat and cook teh astronauts alive from getting too hot. The computers also need cooling, and so are limited by the total ISS cooling budget, which is limited by the size and number of their radiators.
Since the ISS continued longer than planned, they kept bolting on new unplanned additions. things take years to design, test, validate,etc. before flown, and so something planned in the 1980s and initially designed in the 1990s is going to be obsolete 35yrs later, who knew. The ISS was initially developed in the days before personal computers, internet, and other technology being commonplace. Even most electronic calculators in teh early 1990s were still very crude devices, with 2-line calculators and more scientific computing features coming along later.
This video came at just the right time, I've been looking for a server for my space station.
it is ironic that "space age technology" needed to be older and more analog to survive.
And/Or, Because it took so much time to build things and most of the NASA space tech was designed and engineered decades ago, They can't just update/replace to accumulate new tech.. Like for the ISS.
Sounds like the land part of the nuclear triad... which was still using a huge arse diskette...
Not when you consider the "space age" was 60 years ago.
@@mozzjones6943 and yet they can't even follow their own guidelines for making a rocket go figure. joking aside while the getting things to space and having it work issue is a factor most of it's Bureaucracy.
well government has never been good at moving fast. Private industry should be in charge of the tech and integration. I mean a us defense contractror charges 90k dollars for a very small bag of bushings. they charge too much because the government pays with tax dollars.
14:47 for anybody feeling nostalgic, remember AOL and dial-up. You would have to wait 20 minutes for a single song to download. Now you can relive the experience, all you have to do is become an astronaut.
How about sending a >1MB file through MSN? That would also take minutes at best and had a very high potential to fail, at which point you had to start all over again. Sounds very much like what they're dealing with in space today :D
haha, I've been using internet since '89. Downloads to home will slow back then here in Australia. 300 baud modem if I remember correctly?
Wow! This has got to be one of the coolest videos I’ve ever seen on UA-cam! Great to know how the ISS operates and its constraints. It’s nothing like I expected! Please do more videos on space tech if you can 😊
I work in the space industry building rocket engines. Super awesome to see you pump out content that I can relate too on a personal level. Thank you for making this video and I am excitedly waiting now for more space tech content :)
Today's lesson: NASA space station tech doesn't mean high tech, it just means it's engineered to not kill its occupants as far as possible
To be fair, "engineered not to kill its occupants as far as possible" in "lethal high energy high radiation environment" is fairly high tech. :P
To be fair, it is up really high!
It just means "good enough for government work"
Not really surprising.
Think about a submarine - the other kind of pressure vessel we use as a species that deals with dramatic changes. The ISS deals with a total pressure differential of ~1 atmosphere. Relatively speaking, it's easy to keep the thing as one solid piece. The difficult part is the lack of atmosphere.
Take a hinge - like a standard door hinge, two pieces of brass or whatever, that slide together. In space you would find it has the awful habbit of sticking after you open it a few times, and by sticking I mean suddenly welding itself together as the oxide layer is rubbed off, exposing bare smooth metal surfaces that don't know any difference than to presume they are a single continuous piece of metal and suddenly: You can't easily move the hinge. This isn't a problem on earth - any oxide layer that does get removed, quickly reforms. But where you have no atmosphere?
The problems in Space aren't so much about bleeding edge technology - it's about solving problems that you didn't know would be problems. This is why Reliability, trumps basically all else. You need to know it will work.
It's like seeing explosion proof secruity cameras in a companies product line, then thinking they must be some crazy kind of armored camera. Then finding out it just means they won't cause an explosion not that they would survive one.
The very idea of Linus Chaos Monkey Sebastian on the ISS is terrifying.
He would be able to make it drop from orbit.
Where do you think Cobra got the idea for ZEUS during GI Joe 2?
Hijack the communication antenna to get high speed satalite internet 😂
Linus wishes he could go to the ISS so he could drop something in zero gravity and not have to worry about breaking something.
@@TheOmegaRiddler He’d find a way to break it…. This is Linus we are talking about.
Yeah
They launched a hard drive with a super capacitor, but accidentally launched an experiment that tested what happens to unshielded super capacitors when bombarded with gamma radiation. The ISS will ALWAYS manage to gather data about something.
Oh my, I'm curious of the results
@@noobulon4334 The results were simple: SSDs in space fail much faster than on the earth, as the super capacitators break faster and therefore the SSD gets overcharged during saving procedures earlier compared to earth. Solution: regularly backup your data - that's why they have this big server with longer living capacity, so that they can save the data from the experiments all the time and then send it to earth in big packages.
True science is just as excited about a bad result as a good result. Because it's still a result that they can learn from. And let us never forget just how many amazing discoveries were made by accident while trying to learn something entirely different. Accidental results are some of the best we've made.
Thank you so much for the video!! I think this may be my personal favorite. So cool to see the practical applications in such a unique and challenging environment.
Actually found this alot more interesting than I thought I would, great video 😊
Magic School Bus reference was not expected but entirely appreciated. Nice work Editor!
@16:19 "Cryogenic Tanks, do not stir"
Someone at NASA remembers what happened to Apollo 13
Well I would hope that almost everyone at NASA remembere what happened there :D
@@venosaur121212 I hope the exception to that is the kid of the janitor during take your kid to work day, but he would probably still know that.
You missed it at 0:14 then?
@@lmcgregoruk Nope, i just used the one where i was in the video when i decided to make the comment.. :D
they better!
Due to the "lack of gravity", it is impossible to drop something in space.
Linus: hold my LTT screwdriver
You just know he used this opportunity to put a few into the chain there. Hoping that one of them might actually make it up to space and get on one of the live streams they do.
This is so awesome. While watching this, my eyes were sparkling so much, one might have seen the light from the ISS. I love this video. Very well-done.
5:20 missed opportunity in the writing. "In a perfect world..." Well, this deployment is quite literally out of this world.
It's hard to put a computer in space...
unless you put it in a space bar.
Bruh😂
Why did the computer go to space?
To get a byte of the Milky Way!
ba dum tss
god damnit. This joke was from is Dada-base.
@Harsh85114 This is 14 year old bait, if you click don't forget to report! :)
Hi I'm working in aerospace, I can talk about just a couple of things:
1. People think of space as cold, but it might not be as cold as you think. Temperature in low Earth orbit (LEO) can be -60 to +120 C depending on how much direct sunlight you're receiving. Linus' point about having no air to convect heat away is a pretty big deal - that's why you need massive radiator panels.
2. NASA and incumbent space agencies tend to be risk-averse and want to see SIGNIFICANT verification of functionality at every level, particularly when humans are involved. That's probably one reason why they don't use Starlink - the technology is simply too new and maybe there's some technical or bureaucratic inertia to overcome. It's also why tech in space tends to be pretty old, even if it's launched today. US agencies like to see flight heritage. So even (as a silly example) if Ryzen 7600 is better than Intel Core 2, if these agencies see that Intel Core 2 provides everything they need, they're more likely to pick the Core 2 over Ryzen 7600 because they KNOW it'll work. This is compounded by the fact that the rad-hardening process takes a ton of money and time, so you often end up sending surprisingly old tech into space. Newer commercial companies and smallsats are trying to use newer tech though.
3. Power budget is super important. One reason that it's so tight is because solar panels degrade over time, especially in space, so the power available to the ISS is decreasing over time.
4. Bit flips are MUCH more common than on Earth, that's why rad-hardening is important. I had a professor whose research specifically tried to ameliorate the effects of bit-flipping.
5. Not necessarily tech related but it's true that space vehicles need to balance physical practicality versus human preference, e.g. convenient orientations versus "needed" orientations like Linus mentioned.
About Starlink I'm also asking myself, if those satellites even have antennas directing to space. Starlink was built for use on earth primarily.
@@Deinorius They did use Starlink to link to the recent Starship test flight, but that was lower than the satellites. I guess at some point SpaceX will put a dish in/on a Dragon capsule to test it further up.
i work in a company that launches cubesats and we just had a radiation incident a week or so ago! can bus traffic to the power management board completely dead and watchdogs not triggered, radiation’s bad for the little dudes lol
@@Deinorius they have some sort of communication capability in space. I'm not sure off the top of my head what it is - but we know they're expected to be able to respond in case of collisions. Whether that can be repurposed for communication with Earth, I dunno
Why don't they use Radioisotope Thermoelectric Generators instead or in combination with solar? They could reliably generate hundreds of watts for decades.
This brings back memories when i was a test engineer at a satellite company. Nearly every circuit had a duplicate on the same chassis. Then they would put two of those in which is essentially 4 copies. They have to guarantee a satellite to work for 10-15 years without being able to repair it.
That scene from MSB of Arnold taking his helmet off on Pluto, lives rent free in my head. one of these few episodes I ever actually saw (was probably with school tbh)
What an awesome video. I love it that LTT is sticking to the original mission of making informative/educational vids, love it!
The editors had way too much fun with the windows.
Oh yes 😂
i dont get it
@@tomikun8057Look at 10:59
@@tomikun8057 11:00
@@tomikun8057 In the windows of the ISS replica you can see random things like animated UFOs flying by that the editors added in.
1:34 I like that Spaceborne 2 has the same naming scheme as New-New Whonnock
He did call it "New Spaceborn 2" xD
A friend of mine is a physicist who used to work at CERN and, as part of her job, assembled sensors for the particle collider. And despite the sensor's specific purpose being to measure the products of a collision, one of the issues they had to deal with was these particles causing bit flips and other issues when passing through the chips inside the sensors.
So what she told me they did was to irradiate the chips before putting them in because this somehow increased protection from the effects of the particles passing through. Or better put less susceptible to damage. She tried to explain it to me what happens in detail but this was long ago and I didn't understand half of it, so I might be talking BS here. Though I wonder if NASA/ESA also do something similar to their chips.
But I guess this would only be useful for heavier particles than gamma radiation, which I guess is the main problem in space.
All that being said, computing in hostile environments is both very interesting and potentially very frustrating as entropy slaps you in the face.
Physicist here, you pre eradiate them to get those that are particularly susceptible to die before you put it into the heart of a machine you can only service every few years. The procedure does not make anything more resilient, just the surviving units are less likely to fail later on.
@@XanatosDavid Ah okay, that makes much more sense. Then I remembered it wrong. I just remembered something about irradiating chips and some presentation slides she sent me years ago. Though I looked through CERN's DB and found them, and you are completely right, it's just part of a testing campaign. In case you're interested (they mention "irradiation qualification of ASICs"):
indico.cern.ch/event/774201/contributions/3429235/attachments/1875763/3088566/iWoRiDtalkDette2019.pdf
Though since you're here and could provide some background knowledge: Since this is basically just testing/sorting chips, wouldn't this also make sense for any kind of ASIC/CPU/GPU NASA sends up there? Or do they do that already?
@@XanatosDavid Somehow my last response was deleted, probably because I linked to a PDF from the CERN document server with slides my friend sent me back when she told me about this.
But thank you for explaining and of course you are completely right and I remembered it wrong, it was basically just done to sort out the chips that would be most affected by radiation. In the slides she sent me back then she called it "irradiation qualification".
But since I have another physicist to ask now: Do you know if NASA performs similar tests with the computers they send up? Would make a lot of sense I guess.
Oh and in case anyone wonders, the PDF can be found on the CERN document server, it has ID 2686279.
@@B20C0 One of the points of this project was to send off the shelf hardware. Most computers that have been sent up before are treated like so or custom made. The spaceborn project's whole point was "modern enterprise hardware has ECC built into every layer how would it work in space?"
Something like this is actually common practice for all industrial grade (and military and aerospace grade of course) electronics. Most component failures tend to occur early, so the parts undergo a stress test inside a climatic chamber, cycling through the temperature and humidity ranges the parts are specified for. It's called a "burn-in". This is of course expensive, but it's a lot less expensive than fixing a train stuck in some tunnel in the alps or a plane crashing.
Sometimes I see the collabs that LTT does and think "hmm. Strange. Ok though, they're cool." And then I see a collab like this and it's like wtf man where did these connections happen
Fun fact: the computers on board the Perseverance and Curiosity mars sister rovers are 30+ y/o designs chosen because they are radiation hardened, and if you think conputers on the ISS are hard, keep in mind that the farther out you go, the harder it gets
Ok, what happened to the real Linus? This guy at 1:57 is wearing shoes, not sandals
Linus is a synth now
@@aweisen1 I understood that reference.
He probably got told that only closed footwear is allowed on premises...
Must be NASA regulations.
It's ai generated obviously cause hes also not looking 35.
Space is full of computers
More like Space is full of ThinkPads...
Older computers work better in space, because they're less likely to bit-flip.
@@LaughingOrange running arch linux on them makes you cool too
Thats a huge marketing hook lenovo can use but dont, apple would prob pay billions to show an imac in the iss under actual use
@@furinick Only that neither Apple nor NASA would ever attampt to send any Mac-Based system to space. Especially now when there's esentailly no way to swap the OS on Apple Silicon Macs. There is a lot of iPads on the ISS though. Both personal and scientific iPads to track data etc. Funny enough, it was Russia that started to ship iPads to the ISS as enterntainment devices to replace aging iPods in 2011. NASA has since adopted iPads as well though. Sometimes in Crew Dragon launches you can see iPads starpped to the Dragon Flight Suites.
They're actually HP Z Books, and they're old because certifying them for flight is an expensive endeavor, so NASA does a bulk lifetime buy of the laptops and spare parts all at once.
that "Houtson we have a.. Linus" joke was AMAZIIING!!!!!
Hate to buzzkill, but that's based on an oft-repeated misquote. Neither of the Apollo 13 astronauts involved said that. It was closer to "Houston, we've had a...".
Redditor, you have to go back
This is quickly becoming the topgear of tech
This was an awesome video. Saying hi from NASA JPSS Program here. I hope to see more videos like this!
Speaking as someone that helped put about 150 satellites into LEO, in my opinion, space hardening the hardware really only applies when you intend on a super long lifespan - which means you also have no reliable de-orbit plan (so you fully intend on being space debris). For something with a lifespan of under 10 years, you just need some redundant systems and really good error checking in your software. Software based hardening is 100% a thing and way cheaper.
Always depends on the workload, but it is crazy how many workloads can fit in the Rancher SUSE "Cattle not pets" model for services where you build them to deploy as stateless as possible to allow for processes to spin up when needed to process data and crash with no loss if they go down.
Again redundancy, plan for failures to happen, and then you can start to tighten your tolerances to save costs and not just to get it to work at all!
I think I heard somewhere that you can have multiple identical CPUs executing in parallel, that way either you can detect errors and either recalculate that, reset, or if you have 3+ CPUs then you can fix one using the others.
@@LoganDark4357 You might be referring to a TMR (Triple Modular Redundancy) technique.
@@AGuyWhoWantAUsernameI didn't know that had a name . yes!!!!!!
Or when human lives depend on it.
When a satellite mailfunctions you scream, write it off and send a new one. When a person dies you have a lot more questions and issues to deal with. During the "space race" it was kind of a war with russia and some lives lost were acceptable. That's not the case anymore. That's why things are over engineered and thorougly tested. Because there is no triple a or plumber in space.
I think they're actually called "New New New Spaceborn"
Nintendo naming convention.
Spaceborn 2.0
"New!" Spaceborn 2.0 3D
This video is genuinely so interesting! I appreciate the effort you guys went through and I hope to see more videos about industrial computers :)
15:48 "run a computer up HERE" hmmm
Now they can edge in space
Space? Space! Spaaaaaaaaaaaace!
Space. Space. I love space, la la la, I’m in space. Star star, I like stars!
Dad! I'm in space! I'm proud of you, son. Dad, are you space? Yes. Now we are a family again
Space space wanna go to space yes please space. Space space. Go to space.
Getting bored of space.
That’s it? Just space? Where are the balls?
In 1998 a T1 (1.5Mbps) line on earth was considered impressive! Beaming the equivalent out to space was an incredible feat.
i like these educational style vids more than tech reviews
Love the re-use of the SGI logo.
More content like this please? Love the historical and scientific based educational type stuff
The company I work for used to make a peristaltic pump for experiments on the space shuttles. It was made to be smaller, lighter, and more energy efficient than any other pump we made at the time and had to be dumb enough not to rely on anything that radiation could kill.
See THIS is my type of content. Please do more space computer stuff. Hubble control center next? Or maybe an observatory computer?
love this sort of content, infrormative, and not something most channels will get to do, well done!
The Gaben in the glass @1:37 got me good
My grandfather worked on the computers for the Apollo program with ibm he would have loved this video
I started working for a aerospace company last year, i was amazed at how much effort it takes to make a computer work in space!
LEO is pretty easy, below the Van Allen belts. Even laptops work there. GEO is much more fun. Pretty rare to find a processor that works there. I had to design my own but that was 30 years ago.
One of my favorite videos you’ve done.
Great episode Linus, love the theme, would love to see more odd and unique case situations like this! Maybe power plants? Research centres?
12:20 the editors joke was funnier than Linus's "Th-Air" Lol
this video had 2 of my favorite things. Space and Computers.
that picture of an alien got me laughing 😭😭😭
but can it run Crysis?
Deep Frysis
Nope
Linus is the type of person to put a computer in space and play Minecraft
Alright, we NEED more aerospace computer application videos from LTT. This is phenomenal.
lol the office space reference during the white glove test segment 😂
The U.S. government has outsourced everything to private contractors who use research from public colleges and universities, and actual NASA is looking like some kind of retro future aesthetic using off the shelf server parts.
NASA isn't a government agency.
the private sector computers are far more advanced than anything the gov has developed. Making modern computers is not easy, and the US gov doesn't have the people, nor the equipment, nor the innovation to do it themselves. And if they did, without having volumes to sell, it would be ridiculously expensive for a single computer.
NASA has *always* outsourced a ton of stuff to private contractors.
e.g. take the Apollo Guidance Computer (AGC). The MIT Instrumentation Laboratory designed the computer's hardware, OS and UI. Raytheon did the manufacturing.
Fairchild Semiconductor pioneered the first mass-producible ICs in 1960, and the AGC was the first computer to use them. Of course these were still primitive-one flatpack chip roughly 20mm² might contain two 3-input NOR gates, for a total of 6 transistors.
The AGC's memory was even wilder. Its RAM was magnetic core memory (typical for the time) but its ROM was core rope memory. The AGC had 6 core rope modules. Each contained 512 iron/nickel cores (donuts 0.249" in diameter, about 6.3mm) and 192 "sense" wires. For every core, each sense wire either threaded through it (representing 1) or around it (representing 0). Cores had additional wires for address selection and reading out. All of this threading was done by hand-Raytheon had a factory full of needleworkers threading data into the ROM, bit by bit.
They did this because even though core rope memory was incredibly labour-intensive to make, it had unparalleled information density, about 5× that of the magnetic core memory used as RAM. When going into space, mass and volume are paramount.
It also happens to be resilient to bit-flips ;)
A silicon graphics neofetch? Give me the dots, now
yes, the original Spaceborne was started and designed by SGI right as HPE acquired SGI. But was completed and launched after the merger.
You're too young to remember, but the first laptop in space was by Grid Systems. Their product, the Grid Compass 1101, was introduced in 1982 and became notable for its rugged, clamshell design and magnesium case. This model was used by NASA on the Space Shuttle missions, starting with STS-13 in 1983. They were VERY expensive as I recall. The computer featured an Intel 8086 processor, a 320 × 240-pixel electroluminescent display, 340-kilobyte magnetic bubble memory, and a 1200 bit/s modem. Devices such as hard drives and floppy drives could be connected via the IEEE-488 I/O (also known as GPIB or General Purpose Interface Bus). This port made it possible to connect multiple devices to the addressable device bus. It weighed 5 kg (11 lb). The power input is ~110/220 V AC, 47-66 Hz, 75 W.
Word for word copied and pasted from wikipedia.
The 8086 was a ceramic chip.
I appreciate the dad joke to make fun of Linus’ dad joke by the editor at 12:25
This type of video remind me of modern marvel tv show when I was growing up. Love it :)
Idk if you've done a video on Mars rovers or their computers (a quick UA-cam search suggests not 🤷♂️) but the radiation hardening required for those sattelites and computers that go beyond LEO are insane. For instance, the Perseverance rover launched in 2020 has the equivalent of a late-90s PowerPC Mac CPU. The ISS and other LEO sattelites don't need this because they are below the Van Allen radiation belts that capture most of the sun's harmful radiation.
On another note, I don't think the wider public realizes how difficult it can be to change hardware out in space. Every spacewalk takes multiple hours and you only get half of the work done that you could here on the ground. And every upgrade (like adding Starlink antennas) would require a spacewalk.
The work that goes into computers and equipment that goes into space is amazing. Awesome video, love that you could work with NASA and bring light to a wider audience the awesome things they do. 😁👍
Mars helicopter has zero radiation hardening and runs on midrange Snapdragon. LMAO.
Linus is evolving from making videos about cables to making videos about computers on iss💀
"Why don't they just use Starlink? Good question!" In other words, "NASA didn't think about that but now they just might."
It take time to validate HW for space.
not forgetting the possible hacking of any internal systems would be a very good reason to keep the link private,
wich in the case of the iss are almost all vital systems , life support,power, orbital control...ect..
starlink has less coverage than something like iridium which is what we use for our satellites at work, much higher throughput though
They're still on the wait-list. (What's the ZIP code for the ISS's orbit?)
Nowadays, you can use Starlink anywhere on Earth... which doesn't help the ISS. The Polaris Dawn mission is supposed to conduct high-altitude tests of Starlink laser transcievers, which could be more useful.
One of the most interesting video's you guys have done. I enjoyed every second!
Please more space cpu stuff. Loved it and laughed hard at some facts. Thank you!
bruh that''s not atrocious that's like a normal internet connection for me
So in space Linus can't drop anything.....
Linus, you're HUGE(ish) in Japan.
Very impressed with this video. I know Linus knows his shit, but for him to talk this smoothly about the gear and the onboard systems in an engaging and fun manner yet highly informational way takes a fair bit of skill. Well done LTT on this video.
Great sponsorship video. Not every video needs to home/business PC related.
Now see if someone who supplies Cheyenne Mountain will sponsor you and get you access to their super computers. lol
Can’t believe Linus went to the ISS and invented artificial gravity for this video.
Yes but the question everyone is asking is “Can it play Doom!”?
it has a cpu, gpu, and prob some kind of visual output, so i dont know why not?
Because its impossible to throw a PC that far up.
I know this from experience, ahh those were the golden days.
But hey, no gravity, no drops!
(Yes technically some gravitational force)
But it's in free fall all the time, so actually they don't experience any gravitational forces there.
@@zekicay they experience gravitational force. Otherwise they would fly away. But yes. The don't feel any.
I started using Kioxia about 6 years ago for my main hard drive brand in 6 rigs and haven't been let down by them yet for performance, reliability and price.
My ex girlfriend has a very good friend who works at NASA. He is the a main boss. He is in charge of exploratory machines. He's not the coffee guy. We became good friends through an online game she introduced me to. And we've been good friends ever since.
He once told me that they were looking for a chip set that would remain stable and constant no matter how long it was left turned on for. NASA had problems creating one until SONY popped up in one of their lectures... Now most modern spacecraft use the PS1 chipset or an updated version of it. It can be left on indefinitely without overheating or failing.
My favourite space story is the US spent more than a million dollars to make a zero gravity pen for their astronauts and Russia sent their astronauts with a pencil 😂😂😂
I wonder if framework would be a fit for the laptop situation given you could store dozens of boards in a more condensed space than dozens of laptops?
I haven't watched the video yet but I feel an irresistible urge to comment
Watching Arnold freeze his head into an ice cube on Pluto is a frightening core memory.
11:00 I see an LMG in the background, 'Little Man Green' floating by in the view-port.
This is a really cool video, about a topic I wouldn't have thought to investigate on my own. Super cool stuff.
I like these sorts of videos.
I knew building solutions based on a power budget instead of performance was a thing, you just don't see it often.
I know it was a "Sponsored" video, but it had a lot of interesting and informative info! Thanks!
LTT is hands down the most in depth tech video maker on UA-cam.
10:59 I like the little alien floating by the window.
I was wondering the Starlink question before it got conveniently answered. Though maybe it's not worth it if the ISS retirement plans go through?
Gotta appreciate how at 15:03 they made a typo in Computer at the bottom of the welcome screen on the ISS
Having just wrapped up my Master's thesis, which dealt with working within the constraints of edge compute systems for AI in space deployments, this was great timing and a very enjoyable video. Great work on really putting into perspective the constraints one has to work within when dealing with computer systems in space. The university I attended has a lot of projects under NASA funding, so its always nice to hear that NASA's big ISS projects have to deal with many of the same sorts of constraints that we have to put up with on our smaller space platforms.
The Morse code on the Spaceborne Computer 2 welcome page 15:03 says "GREATINGS FROM THE ISS".
Also, the graphic on the left is the Silicon Graphics logo.
Which has Mike, Dave, Carrie, Mark, Rob, and John hidden in it.
I expect one of them can explain all of these things.
The original Spaceborne Computer was made by HPE in 2017, right after they bought SGI, and NASA, SGI, Cray and HP have a long history. I don't know the orgchart but this made me grin.
The kind of content i would’ve been waiting on cable channels to offer a weekly release of
The Apollo 13 reference at 16:17 is amazing.