Hacker News: amarshall

New comment by amarshall in "How we run Firecracker VMs inside EC2 and start browsers in less than 1s"

amarshall — Wed, 17 Jun 2026 22:36:45 +0000

Pretty light on details, heavy on fluff. 9.8s to 3.1s was userfaultd + hugepages, 500ms was PS/2 mouse and… where is the rest of the time to get to 400ms?

New comment by amarshall in "Ask HN: Has anyone replaced Claude/GPT with a local model for daily coding?"

amarshall — Wed, 17 Jun 2026 01:23:00 +0000

No, and having that is not required to know output speed nor the effect of thinking, so I don’t see the point in such a superfluous, indirect question.

As for the question you’re likely asking: benchmarks that include speed across many models and providers available at various places e.g. https://artificialanalysis.ai/leaderboards/models

New comment by amarshall in "Ask HN: Has anyone replaced Claude/GPT with a local model for daily coding?"

amarshall — Mon, 15 Jun 2026 22:37:01 +0000

Thinking doesn’t change output speed. Anthropic’s models are ~ 40–60 t/s median output speed.

New comment by amarshall in "Ask HN: Has anyone replaced Claude/GPT with a local model for daily coding?"

amarshall — Mon, 15 Jun 2026 21:58:27 +0000

Can take the geometric mean of total and active parameters of MoE to get approximate equivalent quality to dense model params. So sqrt(35*10)≈18.7.

The trade-off of MoE is that it is worse but faster for the same total size.

New comment by amarshall in "Ask HN: Has anyone replaced Claude/GPT with a local model for daily coding?"

amarshall — Mon, 15 Jun 2026 21:52:22 +0000

What context length and kv cache quant (if any) are you using? And MTP?

New comment by amarshall in "Linux Terminal Memory Usage"

amarshall — Mon, 11 May 2026 21:56:09 +0000

Probably want `kitty --single-instance` to reduce memory usage when opening multiple.

https://sw.kovidgoyal.net/kitty/invocation/#cmdoption-kitty-...

New comment by amarshall in "Group averages obscure how an individual's brain controls behavior: study"

amarshall — Mon, 04 May 2026 15:08:15 +0000

The actual paper (linked in another comment here) is literally titled “Nonergodicity and Simpson’s paradox in neurocognitive dynamics of cognitive control”. Why they omit it from the blog post version, I’ve no idea.

New comment by amarshall in "Group averages obscure how an individual's brain controls behavior: study"

amarshall — Sun, 03 May 2026 13:12:27 +0000

Seems like a case of Simpson’s Paradox https://en.wikipedia.org/wiki/Simpson%27s_paradox

New comment by amarshall in "Google releases Gemma 4 open models"

amarshall — Thu, 02 Apr 2026 20:25:39 +0000

If you are running on 4090 and get 5 t/s, then you exceeded your VRAM and are offloading to the CPU (or there is some other serious perf. issue)

New comment by amarshall in "Tinybox – A powerful computer for deep learning"

amarshall — Sun, 22 Mar 2026 15:02:21 +0000

Oh I missed the "quad" before 3090.

New comment by amarshall in "Tinybox – A powerful computer for deep learning"

amarshall — Sun, 22 Mar 2026 01:16:57 +0000

You're almost certainly (definitely, in fact) confusing the 120b and 20b models.

New comment by amarshall in "Apple Studio Display and Studio Display XDR"

amarshall — Tue, 03 Mar 2026 19:05:49 +0000

Yes. Even 90 Hz is a noticeable improvement over 60 Hz. I wouldn’t pick it over high-DPI, though.

New comment by amarshall in "Bluetooth Headphone Jacking: A Key to Your Phone [video]"

amarshall — Thu, 01 Jan 2026 17:45:12 +0000

No, the connector is longer than it is tall.

New comment by amarshall in "Bluetooth Headphone Jacking: A Key to Your Phone [video]"

amarshall — Thu, 01 Jan 2026 16:55:49 +0000

Maybe, but Apple doesn’t make them thinner anyway so the argument is invalid. iPhone 6S with headphone jack: 7.1mm thick. iPhone 17 is 7.95mm thick.

New comment by amarshall in "YouTube caught making AI-edits to videos and adding misleading AI summaries"

amarshall — Sat, 06 Dec 2025 11:47:22 +0000

It’s easier than that. Replace shorts with watch in the URL.

New comment by amarshall in "Cognitive and mental health correlates of short-form video use"

amarshall — Thu, 20 Nov 2025 01:13:52 +0000

Non-Shorts can be vertical as well. It works fine, and without the swipe-encouraging UX. If they were “just making long/medium-form videos in portrait now” they don’t need YT Shorts to do that.

New comment by amarshall in "Asus Announces October Availability of ProArt Display 8K PA32KCX"

amarshall — Wed, 05 Nov 2025 05:11:12 +0000

DisplayPort 2.1 (which the monitor supports) provides sufficient bandwidth for 7680x4320@60 Hz 10-bit without DSC when using UHBR20. The press release unfortunately doesn’t clarify whether the monitor supports UHBR20 or only the lower UHBR10 or UHBR13.5 speeds. Of course, the GPU must also support that (Nvidia RTX 5000 only at the moment, as I believe AMD RX 9000 is only UHBR13.5).

New comment by amarshall in "Shai-Hulud malware attack: Tinycolor and over 40 NPM packages compromised"

amarshall — Tue, 16 Sep 2025 21:45:03 +0000

Redownloading everything isn’t a risk when the lock file contains a hash of the download on first update.

New comment by amarshall in "macOS Tahoe"

amarshall — Tue, 16 Sep 2025 13:44:36 +0000

Enabling Reduce Transparency is a sure-fire way to find a dozen bugs within a few hours. It is always quite apparent no one who is empowered tests it at Apple. At least the padding issue is the one Feedback report I sent that got the “more than 10 similar issues” label so it may actual get fixed.

New comment by amarshall in "macOS Tahoe"

amarshall — Tue, 16 Sep 2025 00:27:57 +0000

> SO much padding

No idea on macOS, but turn on Reduce Transparency on iOS and there’s tons of padding most of the time, but then sometimes zero padding. And I mean zero. The edges of buttons and text are at the edge of the underlying background. It’s…embarrassing.