myrmidon

37 ms·

691.
by myrmidon 1y ago
To put some numbers here: Upper body strength and grip strength are around 3 standard deviations apart for men vs women, meaning that an average man is stronger than 99/100 women.
692.
by myrmidon 1y ago
Hm... I'm still not really buying the "turn datacenter off during peak electricity demand" scenario at all, because the ratios just don't seem credible to me: Assuming ~$10M of capex (to buy the datacenter) per MW of ele
693.
by myrmidon 1y ago
Shenzhen I/O begs to disagree :P https://www.zachtronics.com/shenzhen-io/
694.
by myrmidon 1y ago
I do absolutely agree that extrapolating population over more than a few decades is basically a cointoss, but I still thank that exponential growth is far from certain: Basically every industrialized nation has negative population growth
695.
by myrmidon 1y ago
> It will take at least 30 years, probably more like 300. The danger is that we collectively take a more destructive course. I don't think there is a clear trend toward this goal at all. Extrapolating current trends, we are fairly l
696.
by myrmidon 1y ago
I think this example illustrates the biggest problem with carbon credits very clearly: Basically, Meta subsidises all the electricity consumers near that plant to take the blame for the emissions that their datacenters cause. This achieves
697.
by myrmidon 1y ago
You absolutely could do that if you wanted to basically burn money. Datacenters and especially ML training hardware is highly capital intensive and depreciates at basically constant rate regardless of utilization. I see currently no scenari
698.
by myrmidon 1y ago
I call this greenwashing because Meta throws minor amounts of money at an existing plant, then basically claims all the credit for all the CO2-less electricity that is produced there. This is IMO just a CO2 accounting trick, basically
699.
by myrmidon 1y ago
Yes, but even 90% is completely insufficient for a datacenter, and you would have to substitute power for days or weeks (during refueling/maintenance), which makes backup systems unsuitable for the task. Taking this approach would also
700.
by myrmidon 1y ago
This is just PR and carbon greenwashing. The usecase of "cloud provider buys power plant to power data center" does not exist: All current power plants are completely unsuitable as single power source for a data center because upt
701.
by myrmidon 1y ago
This is not actually a density threshold; the more mass you have, the less density you need/get withing the event horizon, meaning bigger black hole => less average mass density ("Schwarzschild density"). And this Schwarzs
702.
by myrmidon 1y ago
Note that the average mass density within the event horizon of a black hole is not particularly large; for big black holes it is actually smaller than the density of main sequence stars (!!). Naively, I would have expected this to provi
703.
by myrmidon 1y ago
Fully agree. Also, using LLMs for things like this can have bad side-effects, too, simply because it raises the noise-floor: By spelling out things that are not noteworthy enough for a human, you make it more difficult to find comments that
704.
by myrmidon 1y ago
> Discounts are fine. Free riding isn’t. Why not? Could you not make a pretty strong argument that avoided negative externalities (CO2 and air pollution from car engines and tires) make public transport worthwhile by itself, and just fun
705.
by myrmidon 1y ago
Sure but those are not really equivalent/comparable in scale; just looking at power/size and conceptual distance from commercial viability, the Chicago pile does not even match up to something like SPARC or JET, much less ITER. A
706.
by myrmidon 1y ago
As an engineer, the difference between "deliberate lies" and "overoptimistic estimates" is often just in the eye of the beholder; Hanlons Razor should be applied IMO. Was ITER overambitious? Timeline and budget unrealist
707.
by myrmidon 1y ago
No. Because flywheels are not limited by material density, they are limited by centrifugal force. High density is actively bad, you want to maximize strength and minimize density for flywheel designs, and this makes you much more likely to
708.
by myrmidon 1y ago
> I call things like ITER "Blazing Saddles" projects. "We have to protect our phony baloney jobs, gentlemen!" I think this is overly harsh and somewhat unfair. You could make the same argument that anything operating
709.
by myrmidon 1y ago
I'm quite skeptical about the improved performance for everyone from earlier segregation; this neglects the negative feedback effects from sticking low performers together I feel, and I dont see how you preserve any semblance of "
710.
by myrmidon 1y ago
Because collective well-being/improvements to the nation as a whole is literally its purpose. Unlike companies, whose purpose is to earn profits for owners and shareholders.
711.
by myrmidon 1y ago
> In Germany you have schools for students targeting university. I was in such a school. Every student there was there to get into university. Yes, but if half the class was not smarter than you (by whatever standard), then that segregat
712.
by myrmidon 1y ago
My view is that schools purpose is teaching everyone not only a common baseline in language/math, but also how to deal with expectations/responsibilities and other people (teachers and classmates). By doing "early segregati
713.
by myrmidon 1y ago
> I had the direct comparison when I went to university. It became very clear that I was learning much more and faster. Sure, but this should be expected. If you filtered out all children with low interest/performance/support i
714.
by myrmidon 1y ago
I can see your point about it being better to switch candidates early, but this is mostly on Biden himself IMO; I can see publicly visible infighting against the current president doing even more damage to the party (long term and immediate
715.
by myrmidon 1y ago
What do you think the democrats did wrong-- what should they have done to win? I'm curious; I currently don't think that election was so easy for them, with having to run a different candidate, and the overton window shifting righ
716.
by myrmidon 1y ago
This doesn't help; e.g. if you buy sneakers made in the US, those might be affected by leather import tariffs, but Amazon is not gonna know by how much and manufacturers are gonna be cagey with exact numbers.
717.
by myrmidon 2y ago
If we were serious about CO2 capture, then the place to start would be big producers (like coal plants): Because that way you need much less scrubbing efficiency and can tolerate much greater overhead while still being effective. If a techn
718.
by myrmidon 2y ago
Very nice, did not know about this! The only problem with vendor extensions like this is that you can't really rely on it, so you're still kinda forced to keep all the (redundant) zero intialization; solving it at the language lev
719.
by myrmidon 2y ago
I would argue that these cases are pretty rare, and you could always get nominal performance with the __noinit hint, but I think this would seldomly even be needed. If you have instances of zero-initialized structs where you set individual
720.
by myrmidon 2y ago
No, just throw the __noinit attribute at every place where its needed. You probably would not even need it in a lot of instances because the compiler would elide lots of dead stores (zeroing) even without hinting.

More ›