Klathra If the AI compute satellite needed to radiate 1 terrawatt of energy and you could run it at 127 C it would need to have 690 km² of surface area, assuming a perfectly black surface for emissivity. As temperature goes up the surface area needed goes down, but even if you could somehow get the electronics to survive much higher temperatures the surface area is still far too large. Am I missing something? Reply
SkyBill40 The efficiency of solar panels would need to make a quantum leap in order to pull off something on this level. That doesn't take into account the other heavy lifting of just getting that mass up there in the first place. And then what? How long before it falls behind newly developed architectures? It's not like it's easy to just send people up there on a mission to hardware swap, especially at some of the orbital levels discussed. I mean… just look at how much work it took to get the HST right. That required multiple shuttle missions to correct. Reply
bit_user vanadiel007 said: Just calculate how many solar panels would be required to generate the amounts of power he's talking about, and you will realize it's a non-starter. I assume the solar panels he'd use would be more efficient, but if we take the example of the ISS, it gets about 100 W/m^2 ( source: https://en.wikipedia.org/wiki/Solar_panels_on_spacecraft#Spacecraft_that_have_used_solar_power ). So, 1 TW would require 1 * 10^10 m^2 or about 10,000 km^2 (i.e. 100×100 km array). I wonder how much energy it would take to shoot that mass into orbit, and what the break-even point would be for offsetting that initial energy expenditure vs. just using ground-based solar. Reply
vanadiel007 bit_user said: I assume the solar panels he'd use would be more efficient, but if we take the example of the ISS, it gets about 100 W/m^2 ( source: https://en.wikipedia.org/wiki/Solar_panels_on_spacecraft#Spacecraft_that_have_used_solar_power ). So, 1 TW would require 1 * 10^10 m^2 or about 10,000 km^2 (i.e. 100×100 km array). I wonder how much energy it would take to shoot that mass into orbit, and what the break-even point would be for offsetting that initial energy expenditure vs. just using ground-based solar. Exactly, which does not even take into account the herculean task of assembling that in space, and ensuring space debris does not destroy it once it's fully "deployed". And while I am far from a space expert, I am thinking such a structure would cast a huge shadow over earth and likely be visible with the naked eye. It would likely be much more practical and less costly to generate that power here on good old earth. Reply
bit_user SkyBill40 said: The efficiency of solar panels would need to make a quantum leap in order to pull off something on this level. Spacecraft typically use more exotic panel technology than terrestrial applications, I think because it's a lot more economical to launch a smaller panel for a given output level, even if it costs more to manufacture. https://en.wikipedia.org/wiki/Solar_panels_on_spacecraft#Types_of_solar_cells_typically_used SkyBill40 said: And then what? How long before it falls behind newly developed architectures? It's not like it's easy to just send people up there on a mission to hardware swap, especially at some of the orbital levels discussed. Oh, that's where his Optimus robots enter the picture! 🤖 SkyBill40 said: I mean… just look at how much work it took to get the HST right. That required multiple shuttle missions to correct. That was also 40 years ago. Most satellites don't need field maintenance. Due to its orbit, it would be virtually impossible to send a manned service mission to JWST, which is far more complex than Hubble. (if this image doesn't load, see it here: https://www.space.com/james-webb-space-telescope-secondary-mirror-deployed ) Reply
bit_user vanadiel007 said: And while I am far from a space expert, I am thinking such a structure would cast a huge shadow over earth and likely be visible with the naked eye. It wouldn't cast a shadow if it's in a polar orbit that aligns with the dusk/dawn boundary. That's what it would have to use for 24/7 sunlight. Otherwise, you'd need to add batteries, which would be a whole lot more mass to launch and something else that needs maintenance/replacement. Reply
Key considerations
- Investor positioning can change fast
- Volatility remains possible near catalysts
- Macro rates and liquidity can dominate flows
Reference reading
- https://www.tomshardware.com/tech-industry/artificial-intelligence/SPONSORED_LINK_URL
- https://www.tomshardware.com/tech-industry/artificial-intelligence/spacex-ceo-elon-musk-says-ai-compute-in-space-will-be-the-lowest-cost-option-in-5-years-but-nvidias-jensen-huang-says-its-a-dream#main
- https://www.tomshardware.com
- MSI MPG X870I Edge Ti Evo Wifi Review: Three M.2’s and smokin’ fast memory support
- How to Unlock Accelerated AI Storage Performance With RDMA for S3-Compatible Storage
- Wall Street warns of rising AI debt risk as stocks slide on wobbly investor confidence — analysts warn of 'systemic risk' as Nvidia share price creaks
- MSI MPG X870I Edge Ti Evo Wifi Review: Three M.2’s and smokin’ fast memory support
- These 16 gadgets under $50 have upgraded my tech life, and most are already on sale for Black Friday
Informational only. No financial advice. Do your own research.