Hacker News: jrk

New comment by jrk in "xAI joins SpaceX"

jrk — Mon, 02 Feb 2026 23:54:06 +0000

The 1TW is the rated peak power output. It's essentially the same in space. The thing that changes is the average fraction of this sustained over time (due to day/night/seasons/atmosphere, or the lack of all of the above).

It's still the same 1TW theoretical peak in space, it's just that you can actually use close to that full capacity all the time, whereas on earth you'd need to over-provision substantially and add storage, so 1TW of panels can only drive perhaps a few hundred GW of average load.

New comment by jrk in "Flux 2 Klein pure C inference"

jrk — Mon, 19 Jan 2026 04:56:23 +0000

The PyTorch version is using the GPU (with Metal Performance Shaders); this C version is currently using (in the docs I saw) a single CPU core, with AMX (via Apple Accelerate BLAS) but not yet with OpenMP for parallelism. It’s not slow because LLM code is bad, but because it’s not running on the same hardware. That said, it’s also not as fast as it is because of the LLM—all the critical code is in kernel libraries it calls (the same as for PyTorch).

New comment by jrk in "Gemini 3 Flash: Frontier intelligence built for speed"

jrk — Thu, 18 Dec 2025 01:47:22 +0000

Why are GPUs necessarily higher latency than TPUs? Both require roughly the same arithmetic intensity and use the same memory technology at roughly the same bandwidth.

New comment by jrk in "HTML as an Accessible Format for Papers (2023)"

jrk — Sat, 06 Dec 2025 16:09:09 +0000

Not sure when you last tried, but Gemini, Claude, and ChatGPT have all supported pretty effective PDF input for quite a while.

New comment by jrk in "AWS is 10x slower than a dedicated server for the same price [video]"

jrk — Wed, 26 Nov 2025 10:01:01 +0000

It is not only not that much more complex, it is often less complex.

Higher-level services like PaaS (Heroku and above) genuinely do abstract a number of details. But EC2 is just renting pseudo-bare computers—they save no complexity, and they add more by being diskless and requiring networked storage (EBS). The main thing they give you is the ability to spin up arbitrarily many more identical instances at a moment’s notice (usually, at least theoretically, though the amount of the time that you actually hit unavailability or shadow quotas is surprisingly high).

New comment by jrk in "AWS is 10x slower than a dedicated server for the same price [video]"

jrk — Wed, 26 Nov 2025 09:52:27 +0000

You know that AWS will come back up. You definitely don’t know whether your own instances will come back or if you’ll need to redeploy it all.

New comment by jrk in "TiDAR: Think in Diffusion, Talk in Autoregression"

jrk — Sat, 22 Nov 2025 21:09:29 +0000

Yes but you can also do the same thing with autoregressive models just by making them smaller. This tradeoff always exists, the question is whether the Pareto curve for diffusion models ever crosses or dominates the best autoregressive option at the same throughput (or quality).

New comment by jrk in "Trying out Gemini 3 Pro with audio transcription and a new pelican benchmark"

jrk — Tue, 18 Nov 2025 23:23:53 +0000

I think they weren’t asking “why can’t Gemini 3, the model, just do good transcription,” they were asking “why can’t Gemini, the API/app, recognize the task as something best solved not by a single generic model call, but by breaking it down into an initial subtask for a specialized ASR model followed by LLM cleanup, automatically, rather than me having to manually break down the task to achieve that result.”

Substrate X-Ray Lithography, a New American Foundry

jrk — Thu, 30 Oct 2025 15:05:45 +0000

Article URL: https://newsletter.semianalysis.com/p/how-to-kill-2-monopolies-with-1-tool

Comments URL: https://news.ycombinator.com/item?id=45760860

Points: 2

# Comments: 0

New comment by jrk in "Diamond Thermal Conductivity: A New Era in Chip Cooling"

jrk — Sat, 25 Oct 2025 13:24:42 +0000

I think the point was not that gem-grade synthetic diamonds are ugly, but that, as industry masters gem-grade production, presumably below-gem-grade production (“ugly synthetic diamonds”) would become cheap enough to deploy in more engineering settings where diamond’s other unique properties were the key concern.

New comment by jrk in "Claude Code on the web"

jrk — Tue, 21 Oct 2025 00:04:54 +0000

This is an established, though advanced, idea.

Sourcegraph Amp (https://sourcegraph.com/amp) has had this exact feature built in for quite a while: "ask the oracle" triggered an O1 Pro sub-agent (now, I believe, GPT-5 High), and searching can be delegated to cheaper, faster, longer-context sub-agents based on Gemini 2.5 Flash.

New comment by jrk in "Nvidia has produced the first Blackwell wafer on US soil"

jrk — Mon, 20 Oct 2025 04:18:43 +0000

It is a few generations behind: Blackwell is still on N4, which is an N5 variant. Meanwhile TSMC has been shipping N3 family processes in large volume products (Apple) for more than 2 years already, and is starting to ramp the next major node family (N2) for Apple et al. next year.

NVIDIA has often lagged on process, since they drive such large dies, but having the first major project demo wafer on N4 now is literally 2 generations behind Taiwan.

New comment by jrk in "Just talk to it – A way of agentic engineering"

jrk — Wed, 15 Oct 2025 15:45:26 +0000

If you go just a few posts back in Peter's own blog he has a video of himself doing exactly this:

https://steipete.me/posts/2025/live-coding-session-building-...

He has posted others over the past few months, but they don't seem to be on his blog currently.

As @simonw mentions in a peer comment, Armin Ronacher also has several great streams (and he's less caffeinated and frenetic than Peter :)

New comment by jrk in "Vibe engineering"

jrk — Wed, 08 Oct 2025 02:47:19 +0000

People of course often do read (and even modify) the model-generated code, but doing so is specifically not “vibe coding” according to the original definition, which was not meant to encompass “any programming with an LLM” but something much more specific: https://simonwillison.net/2025/Mar/19/vibe-coding/

New comment by jrk in "EU court rules nuclear energy is clean energy"

jrk — Fri, 12 Sep 2025 22:18:44 +0000

Many people see top-line rate increases and assume the issue is supply cost, but transmission and distribution have become over 50% of cost everywhere I’ve lived, and are growing fast, regardless of underlying generation or fuel costs. Distribution alone (the neighborhood/local grid) is now roughly matching the supply cost on my MA bill, and though I last lived in CA in 2019, I would be surprised if PG&E weren’t similar.

New comment by jrk in "ReMarkable Paper Pro Move"

jrk — Thu, 04 Sep 2025 03:27:48 +0000

eInk devices are very much not converging to 16:9 or wider aspect ratios. This device is intentionally the size and shape of a reporter's notebook, but there are virtually no other eInk tablets which diverge significantly from more common paper aspect ratios – they all (ReMarkable, Supernote, Boox, Kindle, etc.) are and continue to be exactly what you say you want.

New comment by jrk in "US Intel"

jrk — Tue, 26 Aug 2025 14:52:26 +0000

The OP is talking about fabrication technology, not end products. Even years into their delays getting to 10nm, Intel had more advanced fabrication technology than TSMC until N7 reached volume in 2018.

New comment by jrk in "Intel CEO Letter to Employees"

jrk — Thu, 24 Jul 2025 22:01:22 +0000

Intel’s situation in 2025 is not comparable to the rest of big tech. They have lost technical leadership, bled market share, and started losing a ton of money in a hugely capital-intensive business. They are actually in need of major triage to survive, not just hopping on a belt-tightening trend among still-massively-profitable software companies.

New comment by jrk in "Smollm3: Smol, multilingual, long-context reasoner LLM"

jrk — Wed, 09 Jul 2025 03:38:32 +0000

This is indeed a reasonable cost estimate for competitive short-term H100 rentals (source: much SemiAnalysis coverage, and my own exploration of the market), but there is a critical error (besides the formatting glitch with `*`):

It was 24 days (576 hours) not 24 hours. $663,552 @ $3/hr.

New comment by jrk in "LLM-powered tools amplify developer capabilities rather than replacing them"

jrk — Mon, 21 Apr 2025 16:29:26 +0000

Simon Willison nailed exactly this 2 years ago:

> I've been thinking about generative AI tools as "bicycles for the mind" (to borrow an old Steve Jobs line), but I think "electric bicycles for the mind" might be more appropriate.

> They can accelerate your natural abilities, you have to learn how to use them, they can give you a significant boost that some people might feel is a bit of a cheat, and they're also quite dangerous if you're not careful with them!

https://simonwillison.net/2023/Feb/13/ebikes/