If every university were subject to similar constraints, the average "quality" of research proposals would go down (everybody would have less time to spend on it) but since the pool of research dollars is assumed constant everyone would still get roughly their same slice - just with less overhead.
How it would actually work is only the best schools would keep their funding while lower tier schools would be shut out entirely and be forced to severely reduce their research agendas. There's a school near me that just went from College to University status because they grew their graduate program enough, they would probably not weather the storm the same as MIT.
On a system's level, that's probably the desired outcome in a world where total science funding is shrinking and fewer people can be employed as scientists.
In your example, I'd be more worried about the case where the specialized design reviewer knows what the available sources of grants are and procedure to apply to them, and the professor has since forgotten that knowledge, and so the department now cannot bring in any grants or revenue. That'd kill science even at established institutions like MIT or Yale or Harvard, even if they have very good researchers.
In a proper 2-loop cooling system, the primary loop (with direct electronics contact) and secondary loop (with seawater/external cooling source) are hydraulically isolated by a heat exchanger. The salt water or whatever never gets anywhere near the electronics.
The problem is, it's still in contact with something, even if it's just the secondary loop. Saltwater is not just incredibly aggressive against metal, the major problem with using it for cooling is fouling. Fish, mussels, algae, debris, there are a lot of things that can clog up your entire setup.
Saltwater comes in the air. Just being near it corrodes everything. Both stainless steel and bronze are very expensive. Even if things were made of corrosion proof materials, not everything can be, for strength reasons.
I recently moved all my projects to a self-hosted forgejo instance and have found it quite satisfactory so far. And it's fast! If you're in the market for a github alternative, take a look - there are options.
It’s not fashionable anymore, but I feel that Phabricator deserves an honorable mention as a self-hostable GH alternative too. Actually its “dated” UI is kind of a plus considering how bad everything is now.
I always balked at GitHub, but was impressed with git very soon after I was first introduced; I migrated from an old Gitea instance to Forgejo for my personal projects and have been very happy with it.
I don't really understand the controversy; there are plenty of licenses an author can choose that restricts commercial use of a project. It feels a bit dishonest to release something under a permissive license and then be upset when someone uses your stuff well within the ways you said is perfectly ok.
So many proprietary companies are built on the back of open-source software. Yes, there is no legal responsibility for Warp to donate to Allacritty. But there is a moral obligation. It's not hard to see open-source maintainers and enthusiasts looking at Warp with skepticism. I didn't know that and will be uninstalling Warp, though I stopped using it months ago.
If someone expects to be compensated for their work they should be upfront about it. IMHO it's dishonest/immoral to freely give something away with no expressed expectation of reciprocity and then get upset when someone doesn't reciprocate.
>> If someone expects to be compensated for their work they should be upfront about it.
Definitely and the Alacritty devs have never asked for anything in return for using their software and code. It's mainly others in the community looking at a commercial company forking and then raising $50M and not even contributing. I've seen huge companies, or their higher ups, Github sponsor developers who are building code they use. It's not unheard of.
Power is not the most expensive part of data center lifetime cost; especially these days when you're filling them with several billion dollars of nvidia chips. It's still an important consideration of course, but not the only one.
I don't know if that's really true. Given realistic life cycles of equipment (~10 years, not 3 as commonly believed) the operating power is going to be 75-80% of the TCO, or more.
In fairness your calculation looks at the most expensive element of the DC but ignores all of the associated parts required to utilize the H100: CPU, memory, cooling, etc. No to say that that flips the calculation (I don't have the answer), but it does leave a lot of power out.
Let's be generous and pretend the rest of the hardware is free but double the energy budget of the H100 to account for all of it along with cooling. You're still at only $1k/yr; $10k over 10 years, or 25% of the TCO (ignoring all other costs).
reply