eldenring

5 ms·

1.
by eldenring 2mo ago
Cooling is relatively easy, you just need radiators which are passive, and essentially reduce to a launch cost penalty. You are right that they can't be serviced, but that is missing the point of orbital data centers. The whole point i
2.
by eldenring 2mo ago
Did you read the whole thing? he is clearly talking about having lost focus which is directly in agreement with your take on the Metaverse.
3.
by eldenring 4mo ago
my reply from the other thread fits here too: > this is just not true at all, there are massive leaps from algorithms, data, etc. every year. scale is one axis of many and you need to get them all correct.
4.
by eldenring 4mo ago
this is just not true at all, there are massive leaps from algorithms, data, etc. every year. scale is one axis of many and you need to get them all correct.
5.
by eldenring 4mo ago
2-3x is completely dwarfed by the remaining improvements in training which is still in its infancy relatively
6.
by eldenring 4mo ago
I'm guessing they had a significant revenue spike from gpt 5.4 and gpt 5.5 being so good at coding, and hiccups at anthropic making it easier for programmers to try the models.
7.
by eldenring 4mo ago
Its just not a thing to consider and doesn't happen often.
8.
by eldenring 4mo ago
This article makes 0 sense. Its not up to billing or computer systems or ease of use or anything else that matters. The question is will the scaling laws, which in the asymptote are likely the laws of physics, hold up in converting energy t
9.
by eldenring 5mo ago
CompactStr doesnt have any additional runtime overhead iirc right? So in theory you can drop it in everywhere even when you expect > 25 chars. Maybe an extra branch in the >25 char case?
10.
by eldenring 5mo ago
> Folks are now starting to ask difficult questions about their burn rate and revenue. this view isn't updated correctly post-claude code and codex. there will clearly be sufficient demand.
11.
by eldenring 5mo ago
I don't think so, the whole point of writing software is it is a great sink for complexity. Encoding a process or mechanism in a program makes it work (as defined) for ever perfectly. An example here is in engineering. Building a simul
12.
by eldenring 5mo ago
I think the coding market will be much larger. Knowledge work is kind of like the leaf nodes of the economy where software is the branches. That's to say, making software easier and cheaper to write will cause more and more complexity
13.
by eldenring 5mo ago
Because there's a realistic chance this is the only important software technology moving forward, and commoditizes Metas's entire business which is software.
14.
by eldenring 5mo ago
I don't see how its possible to think this. AI coding assistants are some of the most useful technologies ever created, and model quality is by far the most important thing, so I doesn't make sense why local inference would be the
15.
by eldenring 6mo ago
How many docs do you put in the context? we maintain a lot of dsl code internally, and each file has a copy of the spec + guide as a comment at the top. Its about 50 locs and the relevant models are great at writing it.
16.
by eldenring 6mo ago
I'm not sure what you mean by menial coding but all my employers have supported this in the past. This was a variety of companies, big tech, startups, etc. I think its more likely your employer is the outlier.
17.
by eldenring 7mo ago
Its not subsidized, in fact, they probably have very healthy margins on Claude Code.
18.
by eldenring 7mo ago
The issue is that they have already paid off their datacenter 5x over compared to cloud. For offline, batch training, I don't ses how any amount of risk could offset the savings.
19.
by eldenring 8mo ago
The ideal shape would be a shaded, flat panel perpendicular to the sun right?
20.
by eldenring 8mo ago
This is a common way of thinking. In practice this type of thing is more like optimizing flop allocation. Surely with an infinite compute and parameter budget you could have a better model with more intensive operations. Another thing to co
21.
by eldenring 8mo ago
I've never seen this argument brought up by anyone serious, not in the above post, not in the space datacenter blog by Google, etc. The main benefit is that solar panels go from a complicated mess of batteries + permitting to a very st
22.
by eldenring 8mo ago
> The panels suffer radiation damage they don't suffer on Earth. I don't think this is true, Starlink satellites have an orbital lifetime of 5-7 years, and GPUs themselves are much more sensitive than solar panels for rad damag
23.
by eldenring 8mo ago
What you're describing is one of two mechanisms of shedding heat which is convection, heating up the environment. What the long comment above is describing is a _completely_ different mechanism, radiation, which is __more__ efficient i
24.
by eldenring 8mo ago
Whats your definition for reasonable temp? my envelope math tells me at 82 celsius (right before h100s start to throttle) you'd need about 1.5x the surface area for radiators. Not exactly back to back, but even 3x surface area is reaso
25.
by eldenring 8mo ago
Interesting. According to this other source https://www.satellitetoday.com/connectivity/2026/02/02/space... the filing mentions this > these satellites would operate between 500 km and 2,000 km altitu
26.
by eldenring 8mo ago
Interesting, That could surely simplify things.
27.
by eldenring 8mo ago
Solar in space is a very different energy source in terms of required infrastructure. You don't need batteries, the efficiency is much higher, cooling scales with surface area (radiative cooling doesn't work as well through an atm
28.
by eldenring 8mo ago
Yeah the main benefits are: 1. solar is very efficient at generating energy, no moving parts, simple physics etc. 2. in space you don't deal with weather or daylight cycle, you can just point your panels at the sun and generate very st
29.
by eldenring 8mo ago
The radiators would be lighter compared to the solar panels, and slightly smaller surface area so you can line them back to back I don't think dissipating heat would be an issue at all. The cost of launch I think is the main bottleneck
30.
by eldenring 8mo ago
You can line the solar panels and radiators facing away from each other, and the radiators would take up less surface area. I think maybe the tricky part would be the weight of water + pipes to move heat from the compute to the radiators.

More ›