majorlinux, to Nvidia
@majorlinux@toot.majorshouse.com avatar

Linux gaming is having another big moment. Not a good one, though.

You may want to avoid NVIDIA driver 550 if you're on a laptop

https://www.gamingonlinux.com/2024/06/you-may-want-to-avoid-nvidia-driver-550-if-youre-on-a-laptop/

#Nvidia #Drivers #Linux #laptop #Tech #Update #Hardware #GPU

majorlinux, to Nvidia
@majorlinux@toot.majorshouse.com avatar

Time to patch some holes!

NVIDIA reveal new security issues in their GPU drivers for June 2024

https://www.gamingonlinux.com/2024/06/nvidia-reveal-new-security-issues-in-their-gpu-drivers-for-june-2024/

#Nvidia #Security #GPU #Tech #Hardware

ProjectPhysX, to GraphicsProgramming

#FluidX3D #CFD v2.17 is out! Some huge #GPU/​#CPU hardware has been announced at #Computex, so I've made my code ready. Until now I've been using 32-bit indexing, which overflows for >2³² grid cells in a domain, equivalent to 225 GB VRAM. Now my #OpenCL code will at runtime automatically compile with 64-bit indexing when more cells are used. 🖖🧐
Also, I've added a new raytracing-based field visualization. Thank you @python for the idea! 💡
Release notes 👉 https://github.com/ProjectPhysX/FluidX3D/releases/tag/v2.17

msleigh, to hpc

Not at , but if you are, two members of my team have a mini-symposium coming up: "What the ? Lost in Formula Translation": "...the emergence of GPUs, needing significant refactoring, often beyond simple addition of pragmas, to achieve good performance on established vectorised CPU code, leaves one unspoken question: Why the FORTRAN does my compiler not do this for me?"

https://pasc24.pasc-conference.org/session/?sess=sess127

majorlinux, to apple
@majorlinux@toot.majorshouse.com avatar

We promise you won't miss that 10th core.

Apple Says iPad Air Performance Details Are Accurate Despite 9-Core GPU Mistake

https://www.macrumors.com/2024/06/03/ipad-air-performance-accurate-gpu-error/

kornel, to random
@kornel@mastodon.social avatar

I'm writing a fluid simulator as an excuse to learn more coding.

I get now why people keep reinventing fluid simulators — there's some very clever math in the fundamentals, but also the problem has no right solution, so there's also lot that can be fudged and tweaked for different results.

kornel,
@kornel@mastodon.social avatar

My fluid sim is working!

Debugging of these things is hard, because every step of the sim has a circular dependency on every other step, so a bug in one of them breaks them all.

Now I'm trying to make it a proper plugin and hook up to lights.

ad-hoc visualization of eulerian fluid sim - density, divergence, velocity, pressure

minioctt, (edited ) to GraphicsProgramming Italian

Quei gamer che come me hanno nell’animo il PC masterrace, ma nella pratica odiano il tedio contro cui ci si imbatte appena si vuole effettivamente giocare a qualcosa per PC, staranno con ansia aspettando progressi sulla questione di giochi Windows nel browser. Quindi, che ho fatto ieri? …Un cazzo di niente! 🥰

Ho perso mezza giornata cercando invano di far funzionare prima alcuni programmi di benchmark, e poi Touhou. Ho dovuto installare DirectX 9 per quei programmi, ma poi comunque questi non partivano, perché a quanto pare DX9 non include rendering via software per Direct3D, e in v86 non c’è alcuna GPU. Pazienza, per quelli me ne farò una ragione, non ho vitale bisogno dei benchmark… però per i giochi è una rogna. Sorprendentemente, Touhou 6 parte in questo ambiente mezzo sfasciato, ma c’è un problema: non so come il gioco gira, riesco anche a muovermi nei menu alla cieca e avviare una partita, ma viene renderizzato solo lo sfondo e letteralmente null’altro, quindi non si può effettivamente giocare. (Dettagli più precisi in questa chat qui.) 🤬

https://octospacc.altervista.org/wp-content/uploads/2024/05/photo_5080513907201780824_x8412736725975051660-320x169.jpghttps://octospacc.altervista.org/wp-content/uploads/2024/05/screenshot_2024-05-24-00-21-42-337_org2221662542601539196-960x418.jpgHo provato allora con DirectX 8a, che pare essere l’ultima versione con software rendering prima di DX10 (che lo ha reintrodotto dopo che fu tolto tipo), e stando a Wikipedia ES dovrebbe essere il minimo richiesto da Touhou 6 (altre fonti specificano semplicemente DX8 generico). E qui il gioco però smette di partire, restituendo un errore non molto chiaro… che ho decifrato con Reddit e Google Traduttore, ma la cosa non mi ha dato alcun aiuto, non so a cosa si riferisca l’errore esattamente, e nessuna opzione disponibile tra configurazione del gioco e impostazioni di Windows ha risolto il problema (chiede di cambiare una roba che online non vedo discutere, probabilmente manco esiste). 🐭Proprio stamattina ho voluto dare una chance a Boxedwine; pensandoci, posso usarlo in maniera complementare a v86, nei casi in cui questo non va bene. Fare tutto tra telefono e tablet (dato che non sono a casa) è stato un casino, tra lentezza generale, memoria che si satura (per questo devo usare due dispositivi alla volta) e via… e ugualmente non è servito A NULLA. A parte il tempo perso (tanto, maneggiare con gli ZIP sullo Ximi non è veloce) tentando di avviare gli ultimi pacchetti distribuiti dall’autore per Wine 1.7 e 5.0, che evidentemente non funzionano (il primo non so perché, il secondo è colpa di ), usando il pacchetto di boxedwine.org sono riuscita ad almeno avviare il gioco, ma termina con un errore che credo sia proprio lo stesso di Windows 2000. Nella console di Wine vedo prima del crash una riga fixme:win:EnumDisplayDevicesW ((null),0,0x32f898,0x00000000), stub!, che può confermare i miei sospetti, ma non so se centri o sia una coincidenza. Guardando i file di Boxedwine credo ci sia DirectX dall’8 al 10, quindi non sarà un problema di librerie mancanti. 🐞

Se avete idee, per favore suggerite, perché sono proprio alla frutta, ho finito di giocare ancor prima di iniziare. Il problema è che il gioco non parte con software rendering, anche se in teoria potrebbe usarlo, sospetto che cerchi appunto una GPU e halti se non la trova. (Qui ho caricato il log di Wine.) Ieri sera ho fatto anche tardissimo per andare a letto, per via di questa roba, ma vi giuro ci ho messo un secolo a prendere sonno… questa cosa ora mi tormenta e devo risolverla: so che Touhou è abbastanza leggero da poter girare nel browser, se giochi anche molto più pesanti ce la fanno, quindi ora pretendo di averlo nel browser! Voglio giocarci in Sala Museo! Il problema sono ste cazzo di API; possibile che solo sulle console i giochi funzionano e basta? (Dovrò fare un emulatore Switch per il web, a proposito…) 🧱

https://octospacc.altervista.org/2024/05/24/cant-have-x-in-direct/

gregorni, to Amd
@gregorni@fosstodon.org avatar

I'm not actually hoping for AMD and Intel to overtake NVIDIA in the GPU market. I'd rather see their integrated GPUs get so good that we don't need dedicated GPUs anymore.

#AMD #Intel #NVIDIA #GPU

veronica, to GraphicsProgramming
@veronica@mastodon.online avatar

Is there anyone on here that has a recommendation for NVIDIA graphics cards? I haven't used them for a very long time, only AMD, so not really up to date on them.

I need it to run fine on Linux, which I hope has improved now. I will use it for not very demanding games like WoW, and for Machine Learning. I currently have an AMD Radeon RX 5700 GPU.

I was considering a Dual GeForce RTX 4070 OC, since it would fit in my box. I can't have any of the super long ones.

#GPU #Nvidia #DesktopPC

bornach, to OpenAI
@bornach@masto.ai avatar

All your GPUs are belong to

https://youtu.be/lQNEnVVv4OE
Matthew Berman describes how is gunning for of the
market

lisyarus, to random
@lisyarus@mastodon.gamedev.place avatar

So I was accumulating Monte Carlo samples using good old blending here, now I've switched to a compute shader explicitly blending into a storage texture and it is ~10x slower, interesting

lisyarus,
@lisyarus@mastodon.gamedev.place avatar

So I messed with it a little and it seems that the storage texture itself isn't a problem (writing random values to it works with ~200 fps). Seemingly the core loop over the scene geometry is slower with compute. Interestingly, max ray depth doesn't affect the performance, as if it's prefetching all the data and then looping over it cached.

and folks, does anybody have a clue what's going on? This is wgpu-native, btw.

fell, to Amd
@fell@ma.fellr.net avatar

Apparently, our game crashes with the latest AMD graphics driver, which is version 24.3.1. So I installed 24.3.1 on a testing machine and... it worked fine.

Well apparently AMD made two different "24.3.1" drivers, and the only way to discern them is the article number of the release notes. No seriously. One is the regular RN-RAD-WIN-24-3-1 and the other one is RN-RAD-WIN-24-3-1-POLARIS-VEGA, which is an entirely different driver for the Polaris and Vega cards. :drgn_angry:

fell, to linux
@fell@ma.fellr.net avatar

Always try desktop Linux distributions on real hardware! Do not use virtual machines.

Linux desktops heavily rely on graphics acceleration to deliver a good experience. Most VM tools are meant for servers and do not have good (if any) GPU support.

isagalaev, to ubuntu
@isagalaev@mastodon.social avatar

I'm back! Borked my laptop by upgrading to the recent 24.04 beta. Looks like the upgrade process removed/disabled drivers, and for a while I was left with looking at a plain non-GUI shell, which required an actual real-life magnifying glass, because of my HiDPI display!

Now it's operational again with built-in drivers, but I still want to figure out how to bring the AMD 3-rd part drivers back and restore OpenCL support.

ProjectPhysX, to GraphicsProgramming

update alert! v2.15 speeds up framerate in interactive graphics by 20-70%. 🖖🥳💻

How? Turns out iterating over 2 million pixels with a single CPU core is... really slow. I did that 3 times more than necessary for every frame rendered on screen! 🖖😆
I've now eliminated a memory copy of the frame (in favor of pointer swap), and a clear frame/zbuffer operation on CPU since that's already done on .

Release notes: https://github.com/ProjectPhysX/FluidX3D/releases/tag/v2.15

ErikJonker, to ai
@ErikJonker@mastodon.social avatar

Great article about the hardware behind AI. By the Chairman and Chief Scientist of TSMC who know what they are talking about.
https://spectrum.ieee.org/trillion-transistor-gpu

scott, to GraphicsProgramming
@scott@denizens.social avatar
ProjectPhysX, to GraphicsProgramming

One of my papers got selected for the 2022 Best Paper Award of MDPI Computation! 🖖🥳📃🏆

That was a very bold publication for multiple reasons:

  • I solo-authored it
  • I wrote that paper in only 2 weeks
  • the title contains "Esoteric" twice
  • I submitted it on April 1st

It's serious science though: I discovered a simple algorithm to cut memory demand of the in half, allowing huge simulations on cheap ​s. This is one of the key innovations in .

https://doi.org/10.3390/computation10060092

BenjaminHCCarr, to space
@BenjaminHCCarr@hachyderm.io avatar

missions delayed by shortcomings
NASA's old are causing mission delays — NASA's supercomputers still rely primarily on , and one of its flagship supercomputers uses 18,000 CPUs and 48 . NASA's Office of Inspector General report suggests that transitioning to GPUs and modernization is essential for meeting NASA's current and future needs.
https://www.theregister.com/2024/03/15/nasa_oig_supercomputing_audit

boilingsteam, to linux
@boilingsteam@mastodon.cloud avatar
giuseppebilotta, to GraphicsProgramming

Well, this is interesting.

Someone has posted an announcement for open and positions on the forum
https://gpusph.discourse.group/t/postdoc-ra-positions-at-oregon-state-university/207
Although they are not specifically about GPUSPH the software, they are about , and related topics (including wave modeling for ), so I think I'll leave the announcement up.

wossman, to GraphicsProgramming
@wossman@mastodon.social avatar
mikelundahl, to GraphicsProgramming

Going for something different this time around. Switching out my old 1070 Ti

jay, to ai
ProjectPhysX, to GraphicsProgramming

How realistic can a simulation be? Here is a 1 billion cell simulation of an impacting raindrop, fully raytraced in 8K. FluidX3D contains state-of-the-art volume-of-fluid and surface tension models for highly accurate free surface simulations. Combined with my own engine, results are rendered on-the-fly at resolution as large as remaining VRAM can hold. 🖖😋💧📺
https://youtu.be/MmLNQIW_Sic
FluidX3D is on : https://github.com/ProjectPhysX/FluidX3D

  • All
  • Subscribed
  • Moderated
  • Favorites
  • JUstTest
  • InstantRegret
  • mdbf
  • ethstaker
  • magazineikmin
  • cubers
  • rosin
  • thenastyranch
  • Youngstown
  • osvaldo12
  • slotface
  • khanakhh
  • kavyap
  • DreamBathrooms
  • provamag3
  • Durango
  • everett
  • tacticalgear
  • modclub
  • anitta
  • cisconetworking
  • tester
  • ngwrru68w68
  • GTA5RPClips
  • normalnudes
  • megavids
  • Leos
  • lostlight
  • All magazines