Hacker News: elabajaba

New comment by elabajaba in "Jumping into cold water can stop your heart"

elabajaba — Sat, 25 Apr 2026 17:24:30 +0000

> These were icelanders, around 13 years old, tough as it gets re cold.

Iceland has a famously mild climate, even during winter their average temperature is still above 0C.

New comment by elabajaba in "Amazon is discontinuing Kindle for PC on June 30th"

elabajaba — Sat, 18 Apr 2026 19:27:53 +0000

You do know that both the Kobo Libra and Sage have physical buttons, right?

New comment by elabajaba in "Can I hear a difference between MP3s and uncompressed audio?"

elabajaba — Wed, 25 Mar 2026 08:17:15 +0000

Part of that might be if you're using them wireless because then you're double compressing the audio which amplifies the artifacts (mp3 -> Bluetooth compression).

New comment by elabajaba in "How I learned Vulkan and wrote a small game engine with it (2024)"

elabajaba — Sat, 22 Nov 2025 22:22:14 +0000

Apple officially deprecated GL/GLES on both MacOS and iOS 7 years ago, and only ever supported up to GL 4.1 (which came out in 2010), meaning it doesn't support essential "modern" features like compute shaders (DX11 had them in 2009), or bindless textures (supported since 2012 on AMD+Nvidia, and 2015 for Intel iGPUs, massive performance win, needed for GPU driven rendering and ray tracing).

New comment by elabajaba in "WebGPU and the Price of Compiling WGSL"

elabajaba — Tue, 07 Oct 2025 00:31:25 +0000

No need to transcode to HLSL, DXC already accepts SPIR-V input (and both Chrome and FF are shipping DXC).

New comment by elabajaba in "Unsafe and Unpredictable: My Volvo EX90 Experience"

elabajaba — Tue, 22 Jul 2025 21:52:00 +0000

E-CVTs are extremely reliable and are different from CVTs (CVTs use a belt attached to 2 cones, E-CVTs are just a single planetary gear set), but a lot of car guys and even some mechanics don't realize they're completely different.

New comment by elabajaba in "Alberta separatism push roils Canada"

elabajaba — Sat, 24 May 2025 06:17:08 +0000

Gas car sales peaked in 2018 globally. EVs are already >20% of new car sales worldwide, and the US is a joke when it comes to EV sales compared to Europe or China.

https://ourworldindata.org/electric-car-sales

New comment by elabajaba in "Rust’s dependencies are starting to worry me"

elabajaba — Sat, 10 May 2025 03:27:02 +0000

Part of the rust dependency issue is that the compiler only multithreads at the crate level currently (slowly being improved on nightly, but there's still some bugs before they can roll out the parallel compiler), so most libraries split themselves up into a ton of small crates because otherwise they just take too long to compile.

edit: Also, `cargo-vet` is useful for distributed auditing of crates. There's also `cargo-crev`, but afaik it doesn't have buy in from the megacorps like cargo-vet and last I checked didn't have as many/as consistent reviews.

https://github.com/mozilla/cargo-vet

https://github.com/crev-dev/cargo-crev

New comment by elabajaba in "Migrating away from Rust"

elabajaba — Mon, 28 Apr 2025 23:02:38 +0000

The core unity game engine is c++ that you can't access, but all unity games are written in c#.

New comment by elabajaba in "Framework's first desktop is a strange–but unique–mini ITX gaming PC"

elabajaba — Wed, 26 Feb 2025 01:23:05 +0000

Linux on arm is actually pretty terrible outside of the server space due to their (Qualcomm, Imagination, and ARM) integrated GPUs being bad and having terrible drivers.

New comment by elabajaba in "OpenWrt 24.10.0 – First Stable Release"

elabajaba — Mon, 17 Feb 2025 00:23:42 +0000

R5s has a garbage CPU that won't be able to handle QoS on probably >250mbps.

I'd avoid anything arm based that doesn't have a7x cores (ideally a76/a78 or newer, though I don't think there's any SBC socs using the a710/715/720 yet). A55 cores are old stupidly slow efficiency cores (area efficient, not power efficient).

New comment by elabajaba in "FLAC 1.5 Delivers Multi-Threaded Encoding"

elabajaba — Wed, 12 Feb 2025 00:45:01 +0000

Hell, modern audio codecs (opus and AAC, but not the ffmpeg opus/AAC encoders) are transparent at ~160-192k. MP3 is a legacy codec these days, and generally needs ~30% more bitrate for similar quality.

New comment by elabajaba in "Intel's Battlemage Architecture"

elabajaba — Wed, 12 Feb 2025 00:21:58 +0000

Intel GPU drivers have always been terrible. There's so many features that are just broken if you try to actually use them, on top of just generally being extremely slow.

Hell, the B580 is CPU bottlenecked on everything that isn't a 7800x3d or 9800x3d which is insane for a low-midrange GPU.

New comment by elabajaba in "Intel's Battlemage Architecture"

elabajaba — Tue, 11 Feb 2025 19:27:20 +0000

The amount of memory you can put on a GPU is mainly constrained by the GPU's memory bus width (which is both expensive and power hungry to expand) and the available GDDR chips (generally require 32bits of the bus per chip). We've been using 16Gbit (2GB) chips for awhile, and they're just starting to roll out 24Gbit (3GB) GDDR7 modules, but they're expensive and in limited demand. You also have to account for VRAM being somewhat power hungry (~1.5-2.5w per module under load).

Once you've filled all the slots your only real option is to do a clamshell setup that will double the VRAM capacity by putting chips on the back of the PCB in the same spot as the ones on the front (for timing reasons the traces all have to be the same length). Clamshell designs then need to figure out how to cool those chips on the back (~1.5-2.5w per module depending on speed and if it's GDDR6/6X/7, meaning you could have up to 40w on the back).

Some basic math puts us at 16 modules for a 512 bit bus (only the 5090, have to go back a decade+ to get the last 512bit bus GPU), 12 with 384bit (4090, 7900xtx), or 8 with 256bit (5080, 4080, 7800xt).

A clamshell 5090 with 2GB modules has a max limit of 64GB, or 96GB with (currently expensive and limited) 3GB modules (you'll be able to buy this at some point as the RTX 6000 Blackwell at stupid prices).

HBM can get you higher amounts, but it's extremely expensive to buy (you're competing against H100s, MI300Xs, etc), supply limited (AI hardware companies are buying all of it and want even more), requires a different memory controller (meaning you'll still have to partially redesign the GPU), and requires expensive packaging to assemble it.

New comment by elabajaba in "Intel announces Arc B-series "Battlemage" discrete graphics with Linux support"

elabajaba — Wed, 04 Dec 2024 05:32:53 +0000

To get 128GB of RAM on a GPU you'd need at least a 1024 bit bus. GDDR6x is 16Gbit 32 pins, so you'd need 64 GDDR6x chips, which good luck even trying to fit that around the GPU die since traces need to be the same length, and you want to keep them as short as possible. There's also a good chance you can't run a clamshell setup so you'd have to double the bus width to 2048 because 32 GDDR6x chips would kick off way too much heat to be cooled on the back of a GPU. Such a ridiculous setup would obviously be extremely expensive and would use way too much power.

A more sensible alternative would be going with HBM, except good luck getting any capacity for that since it's all being used for the extremely high margin data center GPUs. HBM is also extremely expensive both in terms of the cost of buying the chips and due to it's advanced packaging requirements.

New comment by elabajaba in "What's Next for WebGPU"

elabajaba — Fri, 22 Nov 2024 03:45:26 +0000

They're talking about the 16 sampled texture binding limit which is the same as webgl2. If you look at eg. the list of devices that are stuck with that few texture bindings they don't even support basic GL with compute shaders or vulkan, so they can't even run webgpu in the first place.

New comment by elabajaba in "Why anti-cheat software utilizes kernel drivers (2020)"

elabajaba — Wed, 30 Oct 2024 23:13:04 +0000

Counterpoint, play any game without anticheat and realize just how much worse it can be. (and before someone says it, custom servers with dedicated admins doesn't scale and tends to cause lots of petty drama)

New comment by elabajaba in "UE5 Nanite in WebGPU"

elabajaba — Fri, 06 Sep 2024 00:24:25 +0000

Intel Arc GPUs are terrible for Nanite rendering, since they lack hardware support for both indirect draws (widely used in GPU driven renderers, Intel emulates it in software which is slow) and 64bit atomics, which are required for nanite.

New comment by elabajaba in "Rust GPU Transitions to Community Ownership"

elabajaba — Fri, 23 Aug 2024 03:34:16 +0000

It looks like they're still actively maintaining a bunch of rust crates, and they're still developing Wim (their blobby Roblox competitor written in rust).

They have pulled back from the rust ecosystem quite a bit though since repi (their former CTO) left shortly after The Finals released. They stopped all their FOSS sponsorships, and there was [this PR](https://github.com/EmbarkStudios/rust-ecosystem/commit/61f0e...) which definitely doesn't inspire confidence.

New comment by elabajaba in "How WASD became the standard PC control scheme (2016)"

elabajaba — Sun, 04 Aug 2024 06:07:02 +0000

Sure, but try hitting 8 without moving your hand from wasd