Ah, I see, thanks.

When I was a child in kindergarten, one time we managed to make two beyblades produce a spark. To this day, that’s one of my proudest moments.

They list unprivileged Docker as the main install method, do you have any references to LXC?

BCD?

I know.

Sure, and that’s why many cloud providers - even ones that don’t train their own models - are only slowly onboarding new customers onto bigger models. Sure. Makes total sense.

Given that cloud providers are desperately trying to get more compute resources, but are limited by chip production - yes, of course? Why do you think they’re trying to expand their resources while their existing resources aren’t already limited?

My guy, we’re not talking about just leaving a model loaded, we’re talking about actual usage in a cloud setting with far more GPUs and users involved.

They are, it’d be uneconomical not to use them fully the whole time. Look up how batching works.

I compared the TDP of an average high-end graphics card with the GPUs required to run big LLMs. Do you disagree?

You can, but that’s not the kind of LLM the meme is talking about. It’s about the big LLMs hosted by large companies.

It’s closer to running 8 high-end video games at once. Sure, from a scale perspective it’s further removed from training, but it’s still fairly expensive.

A “skirt” is an article of clothing, whereas “brisket” is actually pronounced “bruschetta”. You’re welcome!

Fair enough, I hope you’re correct.

Did they explicitly say anywhere that future versions should be coming as well? I only see them mentioning Android 16, and devices still being “supported”, but that could also mean improvements in Android 16.

Could you please link this explanation?

Sadly they might not be able to update GrapheneOS to new Android versions anymore :(

If it’s not a bird, how come it has wings?

Wenn ich draufklicke, bekomme ich folgenden Text:

Das Quorum ist erreicht, wenn 30.000 Mitzeichnungen eingegangen sind. Das Erreichen des Quorums wird in regelmäßigen Abständen, jedoch nicht nach jeder einzelnen Mitzeichnung geprüft. Daher kann es trotz einer angezeigten Mitzeichnerzahl von 30.000 (oder mehr) eine kurze Zeit dauern, bis auch der Status “Quorum erreicht” angezeigt wird.