News for hackers. Stuff that matters.

A comment on How Taalas "prints" LLM onto a chip?.

Re:How Taalas "prints" LLM onto a chip? by Hello9999901 on Sunday February 22, 2026 @01:18AM (#47108704)
This would be a very interesting future. I can imagine Gemma 5 Mini running locally on hardware, or a hard-coded "AI core" like an ALU or media processor that supports particular encoding mechanisms like H.264, AV1, etc.

Other than the obvious costs (but Taalas seems to be bringing back the structured ASIC era so costs shouldn't be that low [1]), I'm curious why this isn't getting much attention from larger companies. Of course, this wouldn't be useful for training models but as the models further improve, I can totally see this inside fully local + ultrafast + ultra efficient processors.

[1] https://en.wikipedia.org/wiki/Structured_ASIC_platform

[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by roncesvalles on Sunday February 22, 2026 @02:28AM (#47109052)
Well even programmable ASICs like Cerebras and Groq give many-multiples speedup over GPUs and the market has hardly reacted at all.
[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by fooker on Sunday February 22, 2026 @04:15AM (#47109570)
> market has hardly reacted at all

Guess who acqui-hired Groq to push this into GPUs?

The name GPU has been an anachronism for a couple of years now.

[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by brainless on Sunday February 22, 2026 @04:28AM (#47109634)
Seems both Nvidia (Groq) and OpenAI (Codex Spark) are now invested in the ASIC route one way or another.
[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by IshKebab on Sunday February 22, 2026 @07:57AM (#47110632)
Cerebras gives a many multiple speedup but it's also many multiples more expensive.
[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by mips_avatar on Sunday February 22, 2026 @08:45AM (#47110951)
The problem with groq was they only allowed LORA on llama 8b and 70b, and you had to have an enterprise contract it wasn't self service.
[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by JKCalhoun on Sunday February 22, 2026 @08:17AM (#47110763)
Apple should have done this yesterday. A local AI on my phone/Macbook is all I really want from this tech.

The cloud-based AI (OpenAI, etc.) are todays AOL.

[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by post-it on Sunday February 22, 2026 @09:23AM (#47111238)
The hardware isn't there yet. Apple's neural engine is neat and has some uses but it just isn't in the same league as Claude right now. We'll get there.
[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by Aurornis on Sunday February 22, 2026 @10:21AM (#47111703)
The die size is huge. This isn’t the kind of chip that would go into your MacBook, let alone an iPhone.

It’s for cloud based servers.

[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by adeelk93 on Sunday February 22, 2026 @11:17AM (#47112215)
And computers used to be the size of a room. I think they can get it to iPhone size in the future, this is an early prototype.
[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by MarsIronPI on Sunday February 22, 2026 @01:01PM (#47113124)
Well, there's a limit to how small we can make transistors with our current technology. As I understand it, Intel is already running into those limits with their new CPUs (they had to redesign the fins IIRC). I can imagine that without an actual breakthrough in chip manufacturing the size could stay large. That's not to say that a breakthrough won't happen, though.
[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by layla5alive on Tuesday February 24, 2026 @12:24AM (#47133176)
Yes, in 2D, but NAND has been using layers for a while. We call HBM interposers 2.5D. 3D breakthrough would be pretty easy but for those pesky problems like power delivery and cooling. (/s)

But give that time (e.g. microfluidics) - something interesting is that it would be extra hard to use all layers at once, but NN might be a good fit, imagining that computation will be sparse (subsets activating simultaneously)...

[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by wmf on Sunday February 22, 2026 @02:04PM (#47113667)
That's the part that people are missing: it won't get smaller. It already required heroic optimization to get 8B on one megachip. Taalas is more expensive but faster. It is cheaper per token when running 24x7 but not cheap to buy. It will never be small and never be cheap.
[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by JKCalhoun on Sunday February 22, 2026 @05:55PM (#47115698)
"It will never be small and never be cheap."

Will your comment age well? We'll see.

We might all be surprised if (somehow, ternary logic?) models come down drastically in size. It doesn't have to be the hardware getting more dense.

[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by fennecbutt on Sunday February 22, 2026 @04:17PM (#47114764)
They did do it yesterday.

And it produced fake headlines and summaries including the threat of lawsuits from involved person(s).

Apple usually waits until somebody else has refined a technology to "invent" it, but I guess they couldn't wait for this one.

[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by wmf on Sunday February 22, 2026 @04:56PM (#47115151)
https://developer.apple.com/documentation/FoundationModels
[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by RobotToaster on Sunday February 22, 2026 @09:18AM (#47111187)
> I'm curious why this isn't getting much attention from larger companies.

I can see two potential reasons:

1) Most of the big players seem convinced that AI is going to continue to improve at the rate it did in 2025, if their assumption is somehow correct by the time any chip entered mass production it would be obsolete.

2) The business model of the big players is to sell expensive subscriptions, and train on and sell the data you give it. Chips that allow for relatively inexpensive offline AI aren't conducive to that.

[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by hrn_frs on Sunday February 22, 2026 @11:49AM (#47112526)
> I'm curious why this isn't getting much attention from larger companies.

Time is money and when you're competing with multiple companies with little margin for error you'll focus all your effort into releasing things quickly.

This chip is "only" a performance boost. It will unlock a lot of potential, but startups can't divide their attention like this. Big companies like google are surely already investigating this venue, but they might lack hardware expertise.

[ Parent | on HN ]
Re:How Taalas "prints" LLM onto a chip? by theptip on Sunday February 22, 2026 @03:09PM (#47114178)
> I'm curious why this isn't getting much attention from larger companies

I would be shocked if Google isn’t working on this right now. They build their own TPUs, this is an extremely obvious direction from there.

(And there are plenty of interesting co-design questions that only the frontier labs can dabble with; Taalas is stuck working around architectural quirks like “top-8 MoE”, Google can just rework the architecture hyperparameters to whatever gets best results in silico.)

[ Parent | on HN ]