ISS today generates and radiates away about 120 KW of energy with its old tech 3250 sq m of panels panels and it's current radiators. That's what 3 H100 racks need
There may be an economic challenge - which seems to be the sort of problem mass manufacturing can solve very well.
There may be a compute model & latency problem, how do you organize model training when racks are much further apart than in traditional data centres (although speed of light is 50% faster in vacuum than glass fibre). But that's algorithms.
Relative to everything else in orbit, powering a rack of compute and some comms per satellite seems not really to be a physics problem.
Well OpenAI is offering equity to the US Government (and who knows who else privately) Tim Apple famously refused to bring manufacturing back to the US when the current president asked and play hardball on infosec. While this is a civil case, increasingly judiciary seems to be an extension of the executive. So it will be interesting to see how this plays out.
Yes, the Supreme Court can reverse anything, and they're re-evaluating the Apple/Fortnite case (which was surprising to a lot of people).
I don't like bringing politics, but the recent 5-4 decision on birthright citizenship doesn't promote a lot of credibility in the current Supreme Court.
Materials science fundamentally has a math & computability problem. Rigorous materials simulations at a nanometre scale seem now feasible - so you could model physical properties of small groups of atoms - eg. Modeling molecular reactions at a sub-atomic level. Bulk property simulations at a centi and up scale also work and so you can do first principles based design and use tricks like finite element methods to deal with the underlying stochasticity to a degree. Materials properties however largely arise from features between nano and micro scale - grain boundaries, dislocations etc. These are computationally infeasible today - there are neither engineering solutions, not math tricks that make this tractable and so materials science and engineering becomes a grind of experimentation and metrology.
This is interesting in that it seems to be making the grind more efficient. I think the true breakthrough will likely be proper scale quantum computing to make the first principles design feasible
Even if you can design a material for specific properties from the ground up, you would still need a way to deliver the kind of microstructure you desire.
Computational design of a target microstructure and material composition is one area, engineering solutions to produce these targets is other level of difficulty.
For those unfamiliar with the reference, in Hitchikers guide to the galaxy series, Mice are projections of hyper intelligent extra terrestrial beings from a higher dimension who commission the creation of planet earth complete with fossils as a giant organic computer
Maybe you just didn't really see it. I've begun to notice that I don't even see certain things that I read, especially if I'm not intellectually prepared to receive them. I vaguely remembered the mice and the earth being a computer (once prompted by the post), but I had no idea about the mice being 3D projections of a multi-dimensional being. I'm pretty sure that idea was either not something I truly comprehended at 20 and/or I thought it was stupid/random (my opinion on the book as a whole), so the thought never even sank in.
Other times I just don't see something for other reasons. I read Augustine's Confessions and never saw the famous line "our hearts are restless until they find their rest in Thee", which is in the THIRD sentence! I'm not sure if the antiquated English [1] threw me off, or what. In fact, I don't remember seeing pretty much every quote people take from Confessions, besides the pear episode. Some of those are theological ideas that are just completely new, some are probably my annoyance at his "confessing" things that aren't really faults [2], and my annoyance at his insistence on quoting the Bible everywhere he could [3]. But I find it strange that I seemed to have simply not seen most of what other people get out of it.
[1] Rant: don't bother reading the out of copyright translations of ancient works, because the 1800s and early 1900s seemed to think that because the works are old, it should be translated into old English. This is stupid. Augustine didn't write in 500 year old Latin, so translating it into 500 year old English is mistranslation.
[2] I am starting to think that maybe "confess" might have meant something different, and it is not confessing a faults but a statement of beliefs.
[3] I suspect that ancient ideas of rhetoric valued quoting trusted sources a lot, kind of how classical Chinese literature quotes older classic literature a lot. Augustine, being a professional rhetorician, would in that case be expected to do that. But it drove me nuts.
Used both. DeepSeek-4 Flash Q2 - last 6 layers Q4 quant with DwarfStar which just about fits in 128Gb is definitely superior IMO - my contexts tend to run typically 50-100k. Throughput tends to be about 12-13k tok/sec - just about acceptable.
Is this advice coming from personal experience of significant diabetes and weight reversal or just online reading informed opinion? Physiology & psychology are quite complex and individualized and "just do this.." doesn't quite work at any sustainable level when people have jobs, families, mortgage, sick children & sick parents to also manage at the same time.
I lost weight by changing my eating habits (which definitely wasn't easy but doable) and have family who used GPT with different results. So a little of both I guess.
I stand by my words that just using Ozembic without changing your habits on the way can't be the right way. And as far as I know studies suggest that that is actually the case. Like with any diet really
You do see this working in many places. Singapore is probably the best example with an explicit scholar programme and competitive pay in much of the upper rungs of the government. This comes with its own problem culturally but does result in very well thought through programs and infra that actually helps people.
Even in the top echelon of the Indian government (Indian Administrative Services), a similar culture exists and has interestingly been strengthend as the power has fragmented across political parties. I personally know several of my top B School/Engineering college mates who joined and any of them would do very well in top management in private industry. You can see the result in the rapidity with which India is pulling people out of poverty and modernizing. Certainly alower than China, but pretty amazing seei living outside for a decade now - many urban services are already far superior in quality to what you get in Europe, US or even Singapore. The challenge in India is the bulk of the bureaucracy below them isn't held to the same standard.In India it's still a prestige thing only in the top echelon - the lower echlons are more motivated by job security, pensions and avenues for corruption.
The difference in Singapore is the government pays well enough to attract and retain higher quality talent deeper.
> We're still in the "$5 airport Uber" era of LLMs. They're heavily subsidized, and everyone still complains about costs.
This is nonsense that AI providers want to peddle. Inference is wildly gross margin profitable - likely 90%+ gross margins. It's very easy to work out the cost structures bottoms up. All providers can drop costs to a third and still keep positive gross margins.
The problems are
1. It possibly still doesn't pay out on training investment in a reasonable time frame without a massive expansion of the 90% gross margin.
2. There is no moat. As we see Mac Mini & High End GPUs stock outs and the pricing offered by DeepSeek and Qwen, the performance of Open Weight models are good enough that people can and are already shifting many inference workloads out of these 90% margin players
Inference I estimate runs 90% plus gross margins. Just work out the math on these servers. I am pretty sure any player can price down. It wouldn't look good on an IPO prospectus.
There may be an economic challenge - which seems to be the sort of problem mass manufacturing can solve very well.
There may be a compute model & latency problem, how do you organize model training when racks are much further apart than in traditional data centres (although speed of light is 50% faster in vacuum than glass fibre). But that's algorithms.
Relative to everything else in orbit, powering a rack of compute and some comms per satellite seems not really to be a physics problem.
reply