By: Andrew Clough (someone.delete@this.somewhere.com), March 24, 2021 8:06 am
Room: Moderated Discussions
RichardC (tich.delete@this.pobox.com) on March 23, 2021 12:47 pm wrote:
> Andrew Clough (someone.delete@this.somewhere.com) on March 22, 2021 4:27 pm wrote:
> > https://millcomputing.com/docs/
>
> They've been banging on about The Mill seemingly forever, without
> ever building any hardware. It's fairly nutty stuff with obvious
> problems handling any branchy code, which is pretty much the definition
> of what a "general-purpose" computer needs to be able to do well.
>
> Back around 2017 they were talking big about having a simulator and
> working on an fpga implementation. As far as I can tell absolutely
> none of that ever happened, at least in public, and it has all gone
> very quiet for 3 years or so.
>
> If it looks like a turkey, and it gobbles like a turkey ...
I agree and disagree.
One the one hand they're trying to innovate on an absurd number of dimensions at once. Some of them are tied together such that they can't be done in isolation but some just seem interesting but gratuitous like some of the cache coherence stuff. Or the branch predictor. Or the funding model! Why would you try out a new startup funding model like that!? So I think they're probably doomed.
But on the other hand they have a lot of very interesting ideas. The videos don't really make clear what is original and what isn't. Like, I hadn't realized exposed pipelines in VLIW had been around for a long time and my first exposure to it was these videos and at the time I thought it was an innovation. But there's tons of stuff in their work that's just really cool to think about and I think looking at their ideas is worth it.
> Andrew Clough (someone.delete@this.somewhere.com) on March 22, 2021 4:27 pm wrote:
> > https://millcomputing.com/docs/
>
> They've been banging on about The Mill seemingly forever, without
> ever building any hardware. It's fairly nutty stuff with obvious
> problems handling any branchy code, which is pretty much the definition
> of what a "general-purpose" computer needs to be able to do well.
>
> Back around 2017 they were talking big about having a simulator and
> working on an fpga implementation. As far as I can tell absolutely
> none of that ever happened, at least in public, and it has all gone
> very quiet for 3 years or so.
>
> If it looks like a turkey, and it gobbles like a turkey ...
I agree and disagree.
One the one hand they're trying to innovate on an absurd number of dimensions at once. Some of them are tied together such that they can't be done in isolation but some just seem interesting but gratuitous like some of the cache coherence stuff. Or the branch predictor. Or the funding model! Why would you try out a new startup funding model like that!? So I think they're probably doomed.
But on the other hand they have a lot of very interesting ideas. The videos don't really make clear what is original and what isn't. Like, I hadn't realized exposed pipelines in VLIW had been around for a long time and my first exposure to it was these videos and at the time I thought it was an innovation. But there's tons of stuff in their work that's just really cool to think about and I think looking at their ideas is worth it.
