genpfault

6 ms·

1.
by genpfault 10d ago
> Gotham Silicon: 1μ CMOS process that will ship custom chips for –$100 in <24hrs > custom chips for –$100 ...are they going to make it up in volume?
2.
by genpfault 22d ago
In particular: https://github.com/n0-computer/dumbpipe https://github.com/cablehead/pai-sho
3.
by genpfault 1mo ago
More generally: https://roguebasin.com/index.php/Articles#Dungeons https://roguebasin.com/index.php/Articles#Map
4.
by genpfault 2mo ago
https://review.lineageos.org/c/LineageOS/android_frameworks_...
5.
by genpfault 2mo ago
> this knockoff site shows SHRDLU[1] has the same thing for $40? Person: Grasp the knockoff. Computer: I DON'T UNDERSTAND WHICH KNOCKOFF YOU MEAN. [1]: https://en.wikipedia.org/wiki/SHRDLU
6.
by genpfault 3mo ago
Not, in fact, optical interferometry :(
7.
by genpfault 3mo ago
What's the app data backup/restore story on GrapheneOS? My understanding is that even with pseudo-D2D (device-to-device) transfers Seedvault doesn't backup everything[1]. Are there more-functional, non-root, local (non-cloud)
8.
by genpfault 3mo ago
C binding: [0] [0]: https://github.com/n0-computer/iroh-c-ffi
9.
by genpfault 3mo ago
> Einstein How are you handling relativistic effects in the N-body simulation?
10.
by genpfault 4mo ago
> KeepassXC can't do auto-type on Wayland https://github.com/keepassxreboot/keepassxc/issues/2281 Though it looks like there's a recent PR for that: https://github.com/keepassxreb
11.
by genpfault 4mo ago
They ever get around to re-adding C++ support?
12.
by genpfault 5mo ago
Prompt processing is absolutely punishing: ./llama-batched-bench -hf unsloth/Qwen3.5-122B-A10B-GGUF:UD-IQ4_NL -npp 1000 -ntg 128 -npl 1 --cache-type-k q8_0 --cache-type-v q8_0 -c 18000 --n-cpu-moe 32 | PP | TG |
13.
by genpfault 5mo ago
Getting ~36-33 tok/s (see the "S_TG t/s" column) on a 24GB Radeon RX 7900 XTX using llama.cpp's Vulkan backend: $ llama-server --version version: 8851 (e365e658f) $ llama-batched-bench -hf unsloth&#x2
14.
by genpfault 5mo ago
Hear, hear!
15.
by genpfault 6mo ago
llama.cpp (b8642) auto-fits ~200k context on this 24GB RX 7900 XTX & it shows a solid 100+ tok/s ("S_TG t/s") on the first 32k of it, nice! ./llama-batched-bench -hf unsloth/gemma-4-26B-A4B-it-GGUF:UD
16.
by genpfault 6mo ago
Doesn't seem to serve rendered samples so you have to set "browser.display.use_document_fonts" to "1" to see anything useful.
17.
by genpfault 6mo ago
600 GB/s of memory bandwidth isn't anything to sneeze at. ~$1000 for the Pro B70, if Microcenter is to be believed: https://www.microcenter.com/product/709007/intel-arc-pro-b70... https://www.
18.
by genpfault 6mo ago
Not to be confused with GNU parallel[1], written in Perl. [1]: https://en.wikipedia.org/wiki/GNU_parallel
19.
by genpfault 6mo ago
Speculative decoding[1]? [1]: https://github.com/ggml-org/llama.cpp/blob/master/docs/specu...
20.
by genpfault 6mo ago
> A social networking system simulates a user using a language model trained using training data generated from user interactions performed by that user Google People[1]? [1]: https://qntm.org/perso
21.
by genpfault 7mo ago
> "A system's purpose is what it does" POSIWID: https://en.wikipedia.org/wiki/The_purpose_of_a_system_is_wha...
22.
by genpfault 7mo ago
https://www.redblobgames.com/articles/curved-paths/
23.
by genpfault 8mo ago
Nifty, thanks for the heads-up!
24.
by genpfault 8mo ago
Nice! Getting ~39 tok/s @ ~60% GPU util. (~170W out of 303W per nvtop). System info: $ ./llama-server --version ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon RX 7900 XTX (RADV NAVI31) (radv) | uma: 0
25.
by genpfault 8mo ago
> ConnectBot https://f-droid.org/en/packages/org.connectbot/ https://github.com/connectbot/connectbot
26.
by genpfault 8mo ago
Because TFA never bothered to define it: Broadband Network Gateway (BNG)[1] [1]: https://github.com/codelaboratoryltd/bng#bng-broadband-netwo...
27.
by genpfault 8mo ago
tap tap tap tap tap
28.
by genpfault 9mo ago
Getting ~150 tok/s on an empty context with a 24 GB 7900XTX via llama.cpp's Vukan backend.
29.
by genpfault 10mo ago
It was right there[1] in the assembly video. [1]: https://youtu.be/pcAEqbYwixU?t=1038
30.
by genpfault 11mo ago
> Ctrl+S to save XOFF ignored, mumble mumble

More ›