MudMan

Not a pop culture thing, it just sounds… self deprecating, I suppose.

Like, was that his legal name? Because man, that’s ballsy.

Anyway, I mostly agree with you, I am more at home here than on all the microblog platforms. I do feel Bsky does a better job at being that than Masto. Masto is insular and man, I hate to say it, but the way its firehose “Home” feed works convinced me that you need at least some options to handle post sorting beyond raw chronological. I don’t hate it, but if I wanted a social network entirely predicated on arguing about Twitter I’d be on Twitter (and for people in denial about the popularity of their open source alternative we have the Linux forums here, so I’m good there, too).

Holy crap, that was a hard pivot. Is your neck OK?

I mean, Bluesky is a private company, it seems to be incorporated in Delaware as a PBC and they claim it is owned by members of Bluesky itself. It’s unclear if that means Dorsey divested from it when he left the board or not, and since they’re not a public company they don’t have an obligation to say. You know as much about Bluesky as you do about Valve or Ikea.

I don’t actively support them or root for them, but it’s not a particularly huge mystery, and my paticipation on it doesn’t imply my moral support to their board members (who are public and known) or their investors.

You can bridge them together and call it a day. Otherwise, presumably because a bunch of the interesting people who left Twitter are there and not on Masto.

I mean, don’t ask questions you don’t want to hear the answers to, I suppose.

The idea is having tensor acceleration built into SoCs for portable devices so they can run models locally on laptops, tablets and phones.

Because, you know, server-side ML model calculations are expensive, so offloading compute to the client makes them cheaper.

But this gen can’t really run anything useful locally so far, as far as I can tell. Most of the demos during the ramp-up to these were thoroughly underwhelming and nowhere near what you get from server-side services.

Of course they could have just called the “NPU” a new GPU feature and make it work closer to how this is run on dedicated GPUs, but I suppose somebody thought that branding this as a separate device was more marketable.

The stupid difference is supposed to be that they have some tensor math accelerators like the ones that have been on GPUs for three generations now. Except they’re small and slow and can barely run anything locally, so if you care about “AI” you’re probably using a dedicated GPU instead of a “NPU”.

And because local AI features have been largely useless, so far there is no software that will, say, take advantage of NPU processing for stuff like image upscaling while using the GPU tensor calculations for in-game raytracing or whatever. You’re not even offloading any workload to the NPU when you’re using your GPU, regardless of what you’re using it for.

For Apple stuff where it’s all integrated it’s probably closer to what you describe, just using the integrated GPU acceleration. I think there are some specific optimizations for the kind of tensor math used in AI as opposed to graphics, but it’s mostly the same thing.