Hacker News: SkyMarshal

New comment by SkyMarshal in "Show HN: Open dataset of real-world LLM performance on Apple Silicon"

SkyMarshal — Thu, 05 Mar 2026 09:28:31 +0000

Nice looking app, downloaded and checking it out. Fyi there's someone else on github keeping track of some LLM benchmarks too:

https://github.com/ggml-org/llama.cpp/discussions/4167

New comment by SkyMarshal in "AMD will bring its “Ryzen AI” processors to standard desktop PCs for first time"

SkyMarshal — Thu, 05 Mar 2026 09:24:55 +0000

Or a Beelink GTR9 Pro with same chip and memory, and two 10GbE LAN ports built in for clustering several together.

New comment by SkyMarshal in "Jensen Huang says Nvidia is pulling back from OpenAI and Anthropic"

SkyMarshal — Thu, 05 Mar 2026 09:21:08 +0000

Why would paying dividends not be like throwing money away for Nvidia, considering the alternative is to reinvest it into Nvidia's R&D, hiring & training, etc. Investors are already happily making money on NVDA stock appreciation, so what more would they gain from paying dividends?

New comment by SkyMarshal in "Loops is a federated, open-source TikTok"

SkyMarshal — Tue, 24 Feb 2026 01:33:53 +0000

https://github.com/joinloops/loops-server/blob/main/INSTALLA...

Looks like an amalgam of php, node, mysql, redis and an AWS S3-compatible filesystem.

New comment by SkyMarshal in "My journey to the microwave alternate timeline"

SkyMarshal — Tue, 24 Feb 2026 01:00:57 +0000

I thought the microwave beeps several times to ensure the radiation has completely dissipated from the chamber before you open it. I always let it beep and then some.

New comment by SkyMarshal in "There's only one Woz, but we can all learn from him"

SkyMarshal — Wed, 28 Jan 2026 09:23:55 +0000

To be fair, Woz wasn't just a “friendly selfless genius inventor engineer”, he was also the co-founder of one of the most valuable tech companies in the world. And YC is, in their own words: "The Y combinator is one of the coolest ideas in computer science. It's also a metaphor for what we do. It's a program that runs programs; we're a company that helps start companies.". They're not entirely unrelated.

A tech company's secretive plan to destroy books

SkyMarshal — Wed, 28 Jan 2026 09:13:17 +0000

Article URL: https://www.msn.com/en-us/news/technology/ar-AA1V4aZv

Comments URL: https://news.ycombinator.com/item?id=46792911

Points: 4

# Comments: 1

New comment by SkyMarshal in "The post-GeForce era: What if Nvidia abandons PC gaming?"

SkyMarshal — Tue, 23 Dec 2025 19:38:25 +0000

I'm not sure it would matter. It doesn't seem that graphics are the limiting factor in games anymore. Plenty of popular games use variations on cartoon-style graphics, for example - Fortnight, Overwatch, Valorant, etc. Seems gameplay, creativity, and player community are more determining factors.

That said, things like improved environmental physics and NPC/enemy AI might enable new and novel game mechanics and creative game design. But that can come from AMD and others too.

There's a war on the bachelor's degree. Can anything replace it?

SkyMarshal — Sun, 26 Oct 2025 15:10:42 +0000

Article URL: https://www.msn.com/en-us/money/careersandeducation/ar-AA1PaQoQ

Comments URL: https://news.ycombinator.com/item?id=45712504

Points: 2

# Comments: 0

New comment by SkyMarshal in "Famous cognitive psychology experiments that failed to replicate"

SkyMarshal — Thu, 18 Sep 2025 04:57:06 +0000

Oblig link to Norvig's "Warning Signs in Experimental Design": https://www.norvig.com/experiment-design.html

New comment by SkyMarshal in "TSMC Arizona: chipmaking is the art of killing variables [video]"

SkyMarshal — Wed, 17 Sep 2025 23:16:32 +0000

This is an interesting insight into the challenges of porting a TSMC fab from Taiwan to Arizona.

There are hundreds of variables that differ, even just based on geography, any one of which can kill nano-scale wafer yields and mean the difference between 90% ("printing money") and 30% ("bleeding cash") yields.

For example, the harder stone bedrock in Arizona has a higher natural resonant frequency than in Taiwan, requiring the fab design to be adjusted to that variable.

Other variables are local air, water, power, suppliers, and culture, all of which differ from Taiwan and require modifications to the original fab design.

TSMC Arizona: chipmaking is the art of killing variables [video]

SkyMarshal — Wed, 17 Sep 2025 23:16:32 +0000

Article URL: https://www.youtube.com/watch?v=1VX3jNJmbcI

Comments URL: https://news.ycombinator.com/item?id=45282607

Points: 2

# Comments: 1

New comment by SkyMarshal in "Mark Zuckerberg freezes AI hiring amid bubble fears"

SkyMarshal — Thu, 21 Aug 2025 21:30:25 +0000

To be fair it worked for him the first time. And for Apple too multiple times for that matter.

New comment by SkyMarshal in "Intel's CEO, Under Attack from Trump, Is Already at Odds with His Board"

SkyMarshal — Tue, 12 Aug 2025 19:18:27 +0000

From another discussion, Intel's board members:

https://news.ycombinator.com/item?id=44842991

Frank Yeary, managing member Darwin capital advisors

James Goetz, partner Sequoia Capital

Andrea Goldsmith, dean of engineering Princeton University

Alyssa Henry, former square CEO

Eric Meurice, former CEO ASML

Barbara Novick, cofounder BlackRock

Steve Sanghi, CEO microchip

Gregory Smith, former CFO Boeing company

Stacy Smith, chair of Autodesk

Dion Weisler, former CEO HP

New comment by SkyMarshal in "A message from Intel CEO Lip-Bu Tan to all company employees"

SkyMarshal — Mon, 11 Aug 2025 23:02:39 +0000

> and they didn't have to spend barely anything on marketing.

To be fair, their marketing was all that gaming related BS you listed, plus developing and maintaining CUDA. Winning gaming got them the winning AI architecture and mindshare around it, and CUDA got them the winning developer interface to it and locked in that mindshare.

New comment by SkyMarshal in "How Boom uses software to accelerate hardware development"

SkyMarshal — Mon, 11 Aug 2025 21:12:22 +0000

I think part of it was that they were testing a new aerodynamic design that eliminates or minimizes sonic boom, so they can go supersonic over land almost immediately after takeoff, and operate over populated land routes. It makes sense to test that kind of thing with the smallest possible model first, then see if you can scale it up to passenger size without losing that quiet acceleration. Their timeline for doing that may be optimistic, but what they're doing makes sense.

New comment by SkyMarshal in "How Boom uses software to accelerate hardware development"

SkyMarshal — Mon, 11 Aug 2025 21:08:11 +0000

That may be true for domestic coast-to-coast flights, but not for transoceanic ones across the Atlantic, or especially the Pacific, or north-south across hemispheres, that can take 8+ hours. Flight time is a higher portion of the total travel time in those cases, and seems like the main market for Boom, especially if they initially target Business Class flyers who do those kinds of trips regularly.

New comment by SkyMarshal in "Ultrathin business card runs a fluid simulation"

SkyMarshal — Fri, 08 Aug 2025 20:20:41 +0000

Just put a QR code on the front that transmits a vCard. Or a way to make the LEDs on the back display a QR code. Then you can still show people your digital business card, even let them hold it and play with it, but it's still obvious the idea is for them to scan the QR code and hand it back.

New comment by SkyMarshal in "GPT-5"

SkyMarshal — Fri, 08 Aug 2025 19:33:22 +0000

> But I think we're not even on the path to creating AGI.

It seems like the LLM model will be component of an eventual AGI, it's voice per se, but not its mind. The mind still requires another innovation or breakthrough we haven't seen yet.

New comment by SkyMarshal in "GPT-5"

SkyMarshal — Fri, 08 Aug 2025 19:30:24 +0000

The inflection point is recursive self-improvement. Once an AI achieves that, and I mean really achieves it - where it can start developing and deploying novel solutions to deep problems that currently bottleneck its own capabilities - that's where one would suddenly leap out in front of the pack and then begin extending its lead. Nobody's there yet though, so their performance is clustering around an asymptotic limit of what LLMs are capable of.