Topology Optimization ANSYS Structural - Search Videos

$A 4B local model can do physics: dynamic fracture, beam mechanics, topology optimization, and Chladni plates. But how do you get a 4B model to "do" physics? We can give it scientific skills that turn intent into experiments, generate data, make images and movies, verify outputs, and report the results. In this video, Gemma-4 E4B runs locally on my MacBook via mistral․rs with 8-bit in-situ quantization, plus the Agent Skills API, completely open-source.Each skill is a folder with a SKILL.md, teac$

A 4B local model can do physics: dynamic fracture, beam mechanics, topology optimization, and Chladni plates. But how do you get a 4B model to "do" physics? We can give it scientific skills that turn intent into experiments, generate data, make images and movies, verify outputs, and report the results. In this video, Gemma-4 E4B runs locally on my MacBook via mistral․rs with 8-bit in-situ quantization, plus the Agent Skills API, completely open-source.Each skill is a folder with a SKILL.md, teac

184 views1 week ago

x.comMarkus J. Buehler

Cool demo. But no, I don't buy that AI made 3D skills optional. Fast output is not the same as a production-ready asset. Topology, cleanup, UVs, rigging, optimization still matter.

Cool demo. But no, I don't buy that AI made 3D skills optional. Fast output is not the same as a production-ready asset. Topology, cleanup, UVs, rigging, optimization still matter.

16.2K views3 weeks ago

x.comRami Sufian

I wouldn't say Tripo meshes have clean topology. Their UV charts also are super scattered. Usually thousands of islands, dozens of triangles each. You need to remesh and generate new UVs to make them usable. Project textures. Generate new normal maps (high->low), etc...

I wouldn't say Tripo meshes have clean topology. Their UV charts also are super scattered. Usually thousands of islands, dozens of triangles each. You need to remesh and generate new UVs to make them usable. Project textures. Generate new normal maps (high->low), etc...

206.2K views2 weeks ago

x.comSebastian Aaltonen

Base mesh creation in under 60 seconds with CozyBlanket Pro's guided AI topology predictor, fully real-time.Currently in development.

Base mesh creation in under 60 seconds with CozyBlanket Pro's guided AI topology predictor, fully real-time.Currently in development.

121.1K views1 week ago

$要約 / Summary400G/800G PAM4 パルストポロジーへの拡張 / Scaling to 400G/800G PAM4 Pulse Topology:4値パルス振幅変調（PAM4）の物理層（00, 01, 10, 11の4電位状態）に対応する、Chiselマルチレーン（4ch/8ch）パラメタライズド・インターフェースを設計。各レーンの到達時間差（スキュー）を「時空のねじれ計量」として処理する、高階（4階）ネットワーク遅延テンソル $\mathcal{L}_{ijk\ell}$ の空間幾何同期回路を構築。可逆計算論に基づく熱散逸の消去 / Elimination of Thermal Dissipation via Reversible Computing:1.50nsの超高速スイッチングに伴う情報消去の熱力学的限界（ランドワーの限界: $\Delta Q = k_B T \ln 2$）を破砕するため、入出力が1:1全単射である「フレドキン（CSWAP）／トフォリ（CCNOT）ゲート」を採用。立方体格子の結晶異方性バグを排したランダム幾何グラフ（RGG）構造上に可逆論理$

要約 / Summary400G/800G PAM4 パルストポロジーへの拡張 / Scaling to 400G/800G PAM4 Pulse Topology:4値パルス振幅変調（PAM4）の物理層（00, 01, 10, 11の4電位状態）に対応する、Chiselマルチレーン（4ch/8ch）パラメタライズド・インターフェースを設計。各レーンの到達時間差（スキュー）を「時空のねじれ計量」として処理する、高階（4階）ネットワーク遅延テンソル $\mathcal{L}_{ijk\ell}$ の空間幾何同期回路を構築。可逆計算論に基づく熱散逸の消去 / Elimination of Thermal Dissipation via Reversible Computing:1.50nsの超高速スイッチングに伴う情報消去の熱力学的限界（ランドワーの限界: $\Delta Q = k_B T \ln 2$）を破砕するため、入出力が1:1全単射である「フレドキン（CSWAP）／トフォリ（CCNOT）ゲート」を採用。立方体格子の結晶異方性バグを排したランダム幾何グラフ（RGG）構造上に可逆論理

1.9K views2 weeks ago

x.comOMUXΩ∞KUT-ASI Junki Kanamori

Updated version online!Same link: https://t.co/UZyj9ViJTw- Performance: ~2× faster path tracing on heavy models (any-hit shadow rays, smarter Russian roulette, GPU-saturating dispatch)- Two-level BVH (TLAS/BLAS) with automatic instance detection- Built-in Khronos sample models (Flight Helmet, chess set, Iridescence Suzanne) — one click to tryStill 100% unbiased — every optimization verified against our white-furnace energy baseline.

Updated version online!Same link: https://t.co/UZyj9ViJTw- Performance: ~2× faster path tracing on heavy models (any-hit shadow rays, smarter Russian roulette, GPU-saturating dispatch)- Two-level BVH (TLAS/BLAS) with automatic instance detection- Built-in Khronos sample models (Flight Helmet, chess set, Iridescence Suzanne) — one click to tryStill 100% unbiased — every optimization verified against our white-furnace energy baseline.

145.1K views2 weeks ago

x.cominvalidUsername.x

3.11 billion Gaussians. I think this is a world record, no?The updates are already sluggish and I can't push further with the hardware limitations I have (mainly disk space). The viewer still sits at around 5GB VRAM.This is a proof of concept that it scales and works. With more optimization there should be no limit. However, I won't drive this forward as it is pretty much pointless as of today.

3.11 billion Gaussians. I think this is a world record, no?The updates are already sluggish and I can't push further with the hardware limitations I have (mainly disk space). The viewer still sits at around 5GB VRAM.This is a proof of concept that it scales and works. With more optimization there should be no limit. However, I won't drive this forward as it is pretty much pointless as of today.

9.6K views2 weeks ago

The joint embedding predictive architecture is flawed itself.You do statistical optimization in embedding space, freeze entire model to use with dynamic programming to determine control commands for, say, a robot.This would'nt generalize as well as he thinks it would.

43.3K views1 week ago

x.comChomba Bupe

$PLTR is the biggest Software Winner in this AI Era 🧵“Once you build a software layer to orchestrate and manage the LLMs in a language your enterprise understands, you actually can create value.” — Dr. Alex Karp, Palantir CEO Full interview on CNBC belowDr. Karp has repeatedly emphasized that standalone large language models and generic prompts fall short in complex enterprise and defense environments. Real, transformative value emerges only through deep implementation: integrating siloed data,

31.2K views2 weeks ago

Morning lads Few updates on this - got the contract for a 10 week period - been already working with the company since April mid- my work is all around multimedia compression pipelines and docker image optimization :p All though not looking to convert this into full time but the team is super chill to work with

60.1K views1 week ago

The SECRET PRO Settings they DON'T TELL YOU 🫵 part #17🚨 BEST WINDOWS SETTINGS for MORE FPS and LOW LATENCYturn on notis, disc in replies, best tools best services.

7.8K views2 weeks ago

x.comSkiioh | PC & Optimization Specialist 🔸

three things happening in AI that nobody's connecting:1. inference costs dropped 1000x in 2 years. not optimization. commoditization. the margin was gone before most companies had a business model.2. RLHF rewards approval, not accuracy. humans rate confident answers higher. so models learned to be confidently wrong in a way that feels right. sycophancy isnt a bug. its the objective function working exactly as designed.3. MoE routes each token to 2-4 specialized sub-networks instead of firing the

5.8M views2 weeks ago

Most AI agents are smart. None of the others are learning.Every session ends and everything the agent figured out disappears. Next session you re-explain everything from scratch.Hermes is the only open-source agent that combines runtime skill creation, persistent multi-layer memory, and an offline optimization pipeline in a single framework.Here is the full architecture behind how it compounds.

23.5K views3 weeks ago

Next gen facial fidelity, fully rigged and production ready. Whether it's for cinematic cutscenes or real time performance, this asset is optimized for fluid, realistic expressions without breaking the topology.Looking to elevate your next project? Let's connect. #UnrealEngine

4.8K views3 weeks ago

x.comGabriel Samuel

Most teams that wanted to deploy AI humans at enterprise scale had to become AI companies first.Tavus Solutions removes that entirely.You bring the workflow. They handle persona creation, conversation architecture, integration, and ongoing optimization.The five pre-built solutions launching today are just the starting point.

124.6K views3 weeks ago

I built a playable 3D roguelike in Unreal Engine 5 from scratch in less than 48 hours.My accelerated workflow:♦ Concepts — Nanobanana Pro♦ 3D Models — Tripo 3D Smart Mesh (under 5-20k polygons, clean topology)♦ Rigging — AccuRig to the UE5 skeleton♦ Weight Painting — Quick Blender cleanup for rigid mesh elements♦ PBR Tile Textures — Patina on https://t.co/w0QJabOcG0♦ Logic — UE5 Blueprint template + ClaudeFull workflow breakdown: https://t.co/mQVRyHyz0C#UE5 #GameDev #3DModeling #Blender3D #TechA

13.9K views1 week ago

x.comStefan 3D AI

🚨 GAMERS DISCORD is STEALING YOUR FPS🚨 BEST DISCORD SETTINGS for MORE FPS, LOWER LATENCY! disc below for the BEST Services, tools, free help & so much more!

11.7K views1 week ago

x.comSkiioh | PC & Optimization Specialist 🔸

so this AI handles SEO, social posts, email outreach, competitor research, lead gen, and GEO optimization…all without any employee?my marketing agency charges me $3k/month and doesn't even do all of that

27.2K views2 weeks ago

See more