Re: https://post.ebin.club/objects/177656be-1e75-4496-b0f9-a473f3d1da65
Uncategorized
1
Posts
1
Posters
0
Views
-
undefined oblomov@sociale.network shared this topic
Gli ultimi otto messaggi ricevuti dalla Federazione
-
Daily Digest | 24 February 2026
Your daily dose of Privacy, Data Protection, AI & Cybersecurity news.
5 stories you should not miss.
Read more: https://www.nicfab.eu/daily-digest/
-
@giuseppebilotta @TomF @glyph i think so. most of the environment is just going to be a small set of models kit bashed with instancing. the instance point sets are to be generated offline (probably in blender) per tile, and then all that's needed is to move the tile relative to the camera's tile. i figure particle systems could work the same way if I end up having any
-
@herrjemineh @activitypub.blog du kannst mal caching probieren. ich habe ganz gute erfahrungen mit surge gemacht... das sollte auch ganz gut mit ActivityPub zusammen spielen...
-
Io sto SEMPRE con il polizziotto (cit.)
-
ENSHITTIFICATION ALERT: START SELF-HOSTING YOUR OWN LLMs.
From: @researchbuzz
https://researchbuzz.masto.host/@researchbuzz/116124902580602857
-
Daily Digest | 24 February 2026
Your daily dose of Privacy, Data Protection, AI & Cybersecurity news.
5 stories you should not miss.
Read more: https://www.nicfab.eu/daily-digest/
-
'When a user asked “What’s the best way to book a weekend away?”, sponsored placements appeared immediately — on the very first response.'
https://searchengineland.com/chatgpt-ads-spotted-and-they-are-quite-aggressive-469651
-
yep, that sounds very similar to what we do in GPUSPH. In our case it's a particle system, and we need an auxiliary grid for binning/neighbors search. The particles are first generated on CPU in world coordinates in double precision, and then positions are split into cell index + single-precision cell fraction (still on CPU), and this is the information that is uploaded and used on the GPU.
Of course in our case the entire particle system is resident on the GPU, so we have the cell index information as a per-particle property on GPU (it's needed to compute the distance to neighboring particles which are in adjacent cells), but if you're only processing one tile at a time in the shader it makes sense to move that to a uniform —if it's needed at all on GPU?
