"If we build these platforms well, we get to ask new questions about what's possible in orbit."
See full article...
See full article...
Investors commit quarter-billion dollars to startup designing “Giga” satellites
- Thread starter JournalBot
- Start date
In a lot of ways, Lin figured she was lucky. She was still cargo, of course, but valuable enough cargo that they allocated enough mass for not only water for the duration of the hop for her, but also food, which was a luxury not many got.
She was going to get to eat during the two days she'd be in orbit. Not many could say that.
Priority was, of course, repairs to the main optilink in the facility; the latest models drew too much power so her first task was to install new solar collectors and repair the couplings to the existing ones. Time was, as always, of the essence as downtime was money and every minute she delayed drove down her own value to the company.
She needed to demonstrate she's worth the food. Hundreds weren't. But she'd prove she was, that she had value, too.
She had 48 hours to repair. To make the drop back to Earth. She didn't want to think on what might happen if she missed her targets. The facility was cold, and her water wouldn't last very long.
Upvote
108
(126
/
-18)
Upvote
43
(51
/
-8)
provides a standard platform for others to build on. So the others don't need to spend time and effort creating their own version of the basics that this provides. The others can specialize in creating the compute or telescope, etc.
Upvote
137
(137
/
0)
Whatever their customer wants to bolt on. Recently, the GEO telecom operators or occasionally a government agency would order a standardized satellite bus like this Boeing 702 and add their own transmitters and receivers. That market is rapidly going extinct, so K2 is making a big bet that new customers will have a business case for big new satellite busses for... AI because that gets the investment money flowing?
But it's not too hard to imagine a second or third generation LEO/MEO with more capability and longer life than Starlink or Starshield. If not commercial customers, defense agencies can always find uses for more satellites...
Upvote
78
(78
/
0)
I see someone misunderstood the joke. Not Herbert Hoover, L. Ron Hoover of the First Church of Appliantology.
Upvote
-15
(8
/
-23)
Gets a $250 million payday for the people who successfully pitched the idea.
Upvote
35
(38
/
-3)
"Space is big. You just won't believe how vastly, hugely, mind-bogglingly big it is. I mean, you may think it's a long way down the road to the chemist's, but that's just peanuts to space."
Upvote
120
(124
/
-4)
I wonder how large this is compared to something like Orion/Mentor which is around 5500kg and 100ish meters when deployed.
Upvote
10
(10
/
0)
Forgive me for my ignorance, could you add or tell me the name of the novel/book/whatever this excerpt comes from? Thanks
Upvote
66
(66
/
0)
Opinions vary.
There seem to be people who will invest in it.
Not your problem, or mine.
Upvote
13
(13
/
0)
Fair point. I wonder, if magnonics ever comes to fruition how it might change things in the thermal management area.
Upvote
-4
(0
/
-4)
In addition, the existing satellite buses are still very, very expensive.
These guys, along with their competitors, are leveraging on the revolution in sat manufacturing. Hundreds of identical sats have been hammered out by multiple producers for a tiny fraction of traditional producer prices. See Iridium onwards.
Upvote
50
(50
/
0)
LEO's carrying capacity is estimated very roughly to be about 100,000 satellites. We're using about 10% of that right now (with production ramping up). Crucially, satellites are routinely de-orbited at the end of their useful life, making room for new satellites.
So... probably, but I guess we'll find out?
Upvote
30
(32
/
-2)
The poster you quoted has sprinkled the Ars comment section with poignant and insightful bits of fiction for a little while at least. Some are more directly topical so I assume they're original creative writing. If so, thanks! And if not, thanks anyways and please share sources!
Upvote
73
(73
/
0)
I get the concept. But I'm curious about the execution. My BS meter is not pleased, but I can't put my finger on exactly what I read that made it squeak louder.
I get the sense that they're offering a product before there's a demand for it. I get the whole, "if you build it, they will come" ideology in space. And I see a lot of VC heading toward space efforts. But as of late, those are to create power stations in space - an endeavor that to me at least sounds stupid on a wide variety of levels. I get that this isn't the same thing, but I also don't see a lot of demand for it yet. At least, demand that has a long-term potential of viability.
Maybe I'm just having a conditioned response to giga this and mega that. Bullshit buzzwords for product names always make me wonder about the quality of the snake oil that's being sold.
I get the sense that they're offering a product before there's a demand for it. I get the whole, "if you build it, they will come" ideology in space. And I see a lot of VC heading toward space efforts. But as of late, those are to create power stations in space - an endeavor that to me at least sounds stupid on a wide variety of levels. I get that this isn't the same thing, but I also don't see a lot of demand for it yet. At least, demand that has a long-term potential of viability.
Maybe I'm just having a conditioned response to giga this and mega that. Bullshit buzzwords for product names always make me wonder about the quality of the snake oil that's being sold.
Upvote
27
(30
/
-3)
Just wrote it off the cuff. Been thinking the last few days about how there seems to be unlimited money for improving AI in the sense of things like orbital datacentres described in the article but never seems to be any for improving the lives of people, and started wondering how long before humans are seen as just another disposable tool, quickly discarded when deemed not cost-effective enough.
Thank you for the kind words!
Upvote
132
(134
/
-2)
Okay, uninformed question here, but whether purposely or otherwise, are there potential issues with Giga-sized satellites de-orbiting? Are they too large to completely burn up on reentry?
Upvote
27
(27
/
0)
Upvote
-4
(2
/
-6)
You'd probably still be able to design things out of thin enough material to ensure destruction. There are some open environmental questions about the wisdom of dumping kilotons of metal and silicon into the upper atmosphere though.
If they're going to GEO, they'll just move to a graveyard orbit when used up and hang out for a few thousand years.
Upvote
39
(39
/
0)
I wonder whether it will ever make sense to "mine" the graveyard orbits for ISRU materials. Maybe after we've been parking dead satellites for a few centuries.
Upvote
29
(30
/
-1)
Just wrote it off the cuff. Been thinking the last few days about how there seems to be unlimited money for improving AI in the sense of things like orbital datacentres described in the article but never seems to be any for improving the lives of people, and started wondering how long before humans are seen as just another disposable tool, quickly discarded when deemed not cost-effective enough.
Thank you for the kind words!
This is not an insult. I found your take very much in the vein of Elysium. Which, I thought, nailed the disposable human angle perfectly.
Great piece of writing. Thank you.
Upvote
22
(23
/
-1)
That sounds like an interesting world to write in. Off the top of my head I could see this turning into some sort of survival adventure with her and one of the satellite AI against the corporation trying to erase them so they don't get in trouble for having one of their AI go rogue.
BTW do you have a place where you put your writings? or is it just personal for now?
Upvote
38
(38
/
0)
I remember reading a couple decades back that in microgravity solder would spontaneously form whiskers that are effectively single molecule crystals at random angles that would inevitably cause short circuits.
Add that to the fact that high energy particles in unshielded space can flip hits in electronics and I’m really not following the logic with these things.
Add that to the fact that high energy particles in unshielded space can flip hits in electronics and I’m really not following the logic with these things.
Upvote
-16
(4
/
-20)
That something can happen is very different than something inevitably happening. I’m pretty sure that there are big GEO satellites that have operated continuously for more than two decades without major impairment of their capabilities. Satellite makers will be constantly learning about potential failure modes and fixing them.
This company intends to use vertical integration and mass production to bring down the cost of satellite types that already exist and allow new types of satellites to be built on the cheaper chassis. They are using the AI hype to raise extra funding, but considering how much easier that makes the fund raising it’s kind of a no brainer decision. And other than that there isn’t really anything special about these satellites except that for some reason nobody else seems to have grabbed the opportunity.
Upvote
42
(42
/
0)
"How do you get rid of the heat?" comes right after "how are you going to power it?"
Upvote
26
(27
/
-1)
Tin whiskers don't happen with leaded solder (the good ol' 63-37 SnPb eutectic alloy) which is why the space industry is basically the last one still allowed to use it. However if you want to leverage commercial components (not the ridiculously expensive, extremely low volume, absurdly mediocre performance "hi-rel"/"space-grade" ones) like I expect all these "newspace" guys do, it does get more and more difficult because notably BGA packages come with their (no-lead) solder balls already attached.
No workable amount of shielding can protect electronics from single-event effects caused by high-energy particles. But as long as these are "only" bit flips and the like, it is reasonably trivial to harden against them through proper software/system design (although the recent memory corruption issue with Airbus A320 elevator control computers shows that it is maybe not so trivial).
All these issues have a strong probabilistic aspects to them however. "Old space" is very risk-averse (and expensive, in a self-reinforcing loop) so they really don't like launching something that has a non-negligible chance of failing. "New space"... not so much
Overall this sounds like "new space" just now (due to expected availability of cheap launch options in the near future) getting their hands at the size of satellite buses that only "old space" used to do (mostly for GEO telecom applications, which have known a massive upheaval due to Starlink). We'll see if the market is here.
Last edited:
Upvote
48
(48
/
0)
ECC is relatively easy to tune in hardware. If you want more protection, add more bits to the ECC.
Yes, that would require custom silicon, but that's just expensive, not difficult. Amazon already has Intel and AMD design them custom chips for their data centers.
If it even turns out to be a problem. I'd test current hardware, and see if it's a problem that needs to be solved.
Upvote
25
(25
/
0)
ECC isn't a panacea. There are still significant issues with excitation events that can't be solved with inline correction systems. Even of earth these issues have caused problems at higher altitude computing centers like Random natlab.
And no, Intel and and aren't doing custom chips for Amazon yet al, they are just selling them slightly different skus of existing COTS designs.
Upvote
14
(15
/
-1)
The way things are going with the commercialization of space, I suspect that these ( and other satellites in graveyard orbits ) will eventually be retrieved and either recycled on-orbit or deorbited.
One thousand years ago steel blades were the latest and greatest human technology. Most humans couldn't read or write and vast swaths of the globe were the unexplored lands of dragons and sea monsters.
One thousand years from now we'll either be extinct, or quite capable of policing up space junk.
Upvote
12
(13
/
-1)