Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

What If Your ESP32 Could Run Real 3D Graphics Without a GPU? 🤯 Jet is a tiny C++17 software 3D rasteriser built for constrained hardware like the ESP32 and STM32. - Runs entirely in software - Integer-only rendering - Native RGB565 - Flat, Gouraud and Phong shading - Affine... show more

Praveen Kumar Verma

7,670 subscribers

52,376 views • 5 days ago •via X (Twitter)

Science & Technology Education

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

THE TOOL EVERY EMBEDDED DEV NEEDS TO KNOW ABOUT no components. no wiring. no setup → wokwi is a browser simulator for arduino, esp32, and stm32 & it actually works drop in sensors, hook up a display, add servos. run your code instantly...everything happens in the browser 1.7M+ projects built by the community, a simon says game, pong on a nano, a 32x32 LED tunnel... all simulated... all shareable learning embedded systems or just need to prototype without touching hardware this is your shortcut

THE TOOL EVERY EMBEDDED DEV NEEDS TO KNOW ABOUT no components. no wiring. no setup → wokwi is a browser simulator for arduino, esp32, and stm32 & it actually works drop in sensors, hook up a display, add servos. run your code instantly...everything happens in the browser 1.7M+ projects built by the community, a simon says game, pong on a nano, a 32x32 LED tunnel... all simulated... all shareable learning embedded systems or just need to prototype without touching hardware this is your shortcut

Vaishnavi

35,148 views • 4 months ago

If you're exhausted from the ambient noise in AI and design, I've got some soup for your soul. Andy Allen has a special point of view about what software could be. He raised $45M, did hardware + software, had a decent exit and walked away from all of it. Now he runs !Boring. Tiny team. No investors. He treats making software like a "kid in the cockpit" with his "hands in the clay". Making a calculator that feels like a game and a camera app that took 3 years to ship. This conversation with Andy was life-giving for me, and I hope you'll enjoy how refreshing his perspective is.

If you're exhausted from the ambient noise in AI and design, I've got some soup for your soul. Andy Allen has a special point of view about what software could be. He raised $45M, did hardware + software, had a decent exit and walked away from all of it. Now he runs !Boring. Tiny team. No investors. He treats making software like a "kid in the cockpit" with his "hands in the clay". Making a calculator that feels like a game and a camera app that took 3 years to ship. This conversation with Andy was life-giving for me, and I hope you'll enjoy how refreshing his perspective is.

Tommy Geoco 🇺🇸

22,117 views • 5 months ago

The quadruped you can assemble yourself. Stella is an educational quadruped robot developed by Ahead. Designed for hands-on learning and self-assembly, the robot uses durable 3D-printed components. Ideal for all education levels, it is easy to repair, infinitely extendable, and equipped with advanced hardware and software. Despite its compact design, Stella is versatile, capable of operating in various environments and even carrying payloads like an optional robotic arm. A powerful tool for exploring robotics.

The quadruped you can assemble yourself. Stella is an educational quadruped robot developed by Ahead. Designed for hands-on learning and self-assembly, the robot uses durable 3D-printed components. Ideal for all education levels, it is easy to repair, infinitely extendable, and equipped with advanced hardware and software. Despite its compact design, Stella is versatile, capable of operating in various environments and even carrying payloads like an optional robotic arm. A powerful tool for exploring robotics.

Circuit

23,215 views • 1 year ago

What If Your Desk Robot Could Learn Your Moves? Meet Quaddle, a tiny open-source quadruped built for tinkering. It has just 4 servos, but can run, glide sideways and spin in place. You can program it with Block coding, Python or C++. You can physically guide it through a movement, tap to record, and it can learn and replay the trick. There is also AI conversation, touch interaction, sensors and 3D-printable parts. A fun way to get hands-on with robotics, coding, AI and mechanical design.

What If Your Desk Robot Could Learn Your Moves? Meet Quaddle, a tiny open-source quadruped built for tinkering. It has just 4 servos, but can run, glide sideways and spin in place. You can program it with Block coding, Python or C++. You can physically guide it through a movement, tap to record, and it can learn and replay the trick. There is also AI conversation, touch interaction, sensors and 3D-printable parts. A fun way to get hands-on with robotics, coding, AI and mechanical design.

Praveen Kumar Verma

14,776 views • 13 days ago

Apple Maps 3d is insanely impressive now (see video below) Google on the other hand decided to build an insanely clunky solution to display 3d maps to avoid using mobile GPUs: A Google server streams (!) a live video of its 3d rendered map to you, the result is a disastrous: a super laggy and clunky 3d view full of video compression artefacts with a response time of 2-3 seconds for every movement And I know WHY they built that, someone felt it was not inclusive to only have 3d maps that are high performing on iOS because Apple's GPUs are insanely fast and can deal with it (as you can see below in Apple Maps) but they wanted to support inferior phones, I get it But then they also decided to worsen the experience for iOS users who could easily run 3d natively on its GPU It's just an insane product decision, there's no way to cut it: they could have super smooth 3d like Apple Maps but neutered it with a clunky live stream Even more insane because Google does have native 3d on iOS in a completely different app that nobody uses: Google Earth Is 3d important? Not rly for nav no but it shows again something about Google's engineering decisions, like the redesign of the Google OAuth Login screen that made zero sense Nobody is speaking up against bad ideas in Google meetings, everyone's too scared of HR maybe, it's the opposite of a meritocracy and it's just super visible from using Google products right now

Apple Maps 3d is insanely impressive now (see video below) Google on the other hand decided to build an insanely clunky solution to display 3d maps to avoid using mobile GPUs: A Google server streams (!) a live video of its 3d rendered map to you, the result is a disastrous: a super laggy and clunky 3d view full of video compression artefacts with a response time of 2-3 seconds for every movement And I know WHY they built that, someone felt it was not inclusive to only have 3d maps that are high performing on iOS because Apple's GPUs are insanely fast and can deal with it (as you can see below in Apple Maps) but they wanted to support inferior phones, I get it But then they also decided to worsen the experience for iOS users who could easily run 3d natively on its GPU It's just an insane product decision, there's no way to cut it: they could have super smooth 3d like Apple Maps but neutered it with a clunky live stream Even more insane because Google does have native 3d on iOS in a completely different app that nobody uses: Google Earth Is 3d important? Not rly for nav no but it shows again something about Google's engineering decisions, like the redesign of the Google OAuth Login screen that made zero sense Nobody is speaking up against bad ideas in Google meetings, everyone's too scared of HR maybe, it's the opposite of a meritocracy and it's just super visible from using Google products right now

@levelsio

1,488,981 views • 2 years ago

Paranoid man mocked for hunting a spy device in his home. No one believed the government could plant bugs in people’s apartments. A month later, his device found a bug in every mocker’s house. He started selling it and made $255k as everyone rushed to check their own homes. Here's how he made this technology using only Claude: The device is called an RF bug detector. Hidden cameras, microphones, and GPS trackers all leak radio signals when they transmit. This scanner picks up those signals and shows you exactly where they are coming from. The whole build costs around $25 in parts and runs on a small rechargeable battery. You need 4 things. - ESP32 development board - AD8317 RF detector module (the part that senses the signal) - A small OLED screen - A LiPo battery and charging board 1. Wire the AD8317 module to the ESP32 analog input, then connect the OLED screen to the I2C pins, which takes about ten minutes on a breadboard. 2. Plug the ESP32 into your computer with a USB cable. 3. Open Claude Code and paste this prompt: { Write me ESP32 firmware for an RF bug detector. Hardware: - AD8317 logarithmic RF detector on analog pin 34 - SSD1306 OLED screen on I2C pins 21 and 22 - LiPo battery powered Requirements: 1. Read the AD8317 voltage 20 times per second. 2. Convert the voltage into a signal strength value in dBm. 3. Show a live bar graph on the OLED that rises as signal gets stronger. 4. Play a beep through a buzzer on pin 25 that speeds up as the signal increases, so I can sweep a room and home in on the source. 5. Add a calibration button on pin 4 that sets the current room level as the baseline, so only stronger signals trigger alerts. Output one complete Arduino sketch with comments so I can tune the sensitivity. } 4. Flash the firmware to the ESP32 straight from the Arduino IDE. 5. Walk slowly around any room holding the device, and watch the bar graph climb as you get closer to a hidden transmitter. 6. Once it works, drop the whole thing into a 3D-printed handheld case so it looks like a finished product. The man in the hook sweeps a room in under two minutes. The beep speeds up as he gets closer, so he can pinpoint a bug hidden inside a smoke detector, a plug, or a lamp. His parts cost around $25 per unit. He sells the finished detector for $129 to renters, travelers, and Airbnb guests who want to check their space. The market is everyone who ever walked into a hotel room and wondered if someone was watching.

Paranoid man mocked for hunting a spy device in his home. No one believed the government could plant bugs in people’s apartments. A month later, his device found a bug in every mocker’s house. He started selling it and made $255k as everyone rushed to check their own homes. Here's how he made this technology using only Claude: The device is called an RF bug detector. Hidden cameras, microphones, and GPS trackers all leak radio signals when they transmit. This scanner picks up those signals and shows you exactly where they are coming from. The whole build costs around $25 in parts and runs on a small rechargeable battery. You need 4 things. - ESP32 development board - AD8317 RF detector module (the part that senses the signal) - A small OLED screen - A LiPo battery and charging board 1. Wire the AD8317 module to the ESP32 analog input, then connect the OLED screen to the I2C pins, which takes about ten minutes on a breadboard. 2. Plug the ESP32 into your computer with a USB cable. 3. Open Claude Code and paste this prompt: { Write me ESP32 firmware for an RF bug detector. Hardware: - AD8317 logarithmic RF detector on analog pin 34 - SSD1306 OLED screen on I2C pins 21 and 22 - LiPo battery powered Requirements: 1. Read the AD8317 voltage 20 times per second. 2. Convert the voltage into a signal strength value in dBm. 3. Show a live bar graph on the OLED that rises as signal gets stronger. 4. Play a beep through a buzzer on pin 25 that speeds up as the signal increases, so I can sweep a room and home in on the source. 5. Add a calibration button on pin 4 that sets the current room level as the baseline, so only stronger signals trigger alerts. Output one complete Arduino sketch with comments so I can tune the sensitivity. } 4. Flash the firmware to the ESP32 straight from the Arduino IDE. 5. Walk slowly around any room holding the device, and watch the bar graph climb as you get closer to a hidden transmitter. 6. Once it works, drop the whole thing into a 3D-printed handheld case so it looks like a finished product. The man in the hook sweeps a room in under two minutes. The beep speeds up as he gets closer, so he can pinpoint a bug hidden inside a smoke detector, a plug, or a lamp. His parts cost around $25 per unit. He sells the finished detector for $129 to renters, travelers, and Airbnb guests who want to check their space. The market is everyone who ever walked into a hotel room and wondered if someone was watching.

Coin Shot ☁️

67,031 views • 2 months ago

Here is the Geometry Nodes Weighted Normals with Laplacian Blur on a full character (a vroid). It easily improves the shading even on game topology with almost no setup. I built this as part of my quest to improve real time toon shading. 3D anime models are popular, but use of dynamic light is rare even among high quality vtuber models. This is for several reasons, but a big one is simply that it takes a lot of Custom Normals work to make 3D cel shading not look like a jagged mess (other pieces of the puzzle are issues like deformations, multiple lights, etc). And fixing Normals is tedious, especially on existing game topology. I have focused on proxy meshes for priority areas like character faces, but they aren't an efficient solution for the whole body + outfit. I wanted something I could just throw on any model and make it at least not a jagged mess anymore even if it wasn't perfect. As you can see from this clip, this does that very well! And vertex groups can be used to control the style of the effect and power. It still can't smooth beyond what the topology density can support, but the topo itself is no longer a problem (for higher res, could be run on a subdivided version of the model and then baked to a Normal Map.) The only changes I made to this model were adding a weld modifier to merge split edges during interpolation, and a vertex group to select the skirt. I have not yet added full handling and logic for detecting edges with big angles like the skirt, or for handling boundaries like on the hair, so both those areas can get better too. You can also see that while it successfully smooths out the Face, it isn't really stylistically correct there. That is still best done with a proxy mesh to define a new shape. This is part of the tools I am working on for Fondant. We are putting together a Blender Addon to release this + a proxy mesh tool for the face, and are working on resolving other problems in-engine to fully bring dynamic light to real time 3D toon shading. Give us a follow, and send them a DM if you are interested in testing these tools as they develop!

Here is the Geometry Nodes Weighted Normals with Laplacian Blur on a full character (a vroid). It easily improves the shading even on game topology with almost no setup. I built this as part of my quest to improve real time toon shading. 3D anime models are popular, but use of dynamic light is rare even among high quality vtuber models. This is for several reasons, but a big one is simply that it takes a lot of Custom Normals work to make 3D cel shading not look like a jagged mess (other pieces of the puzzle are issues like deformations, multiple lights, etc). And fixing Normals is tedious, especially on existing game topology. I have focused on proxy meshes for priority areas like character faces, but they aren't an efficient solution for the whole body + outfit. I wanted something I could just throw on any model and make it at least not a jagged mess anymore even if it wasn't perfect. As you can see from this clip, this does that very well! And vertex groups can be used to control the style of the effect and power. It still can't smooth beyond what the topology density can support, but the topo itself is no longer a problem (for higher res, could be run on a subdivided version of the model and then baked to a Normal Map.) The only changes I made to this model were adding a weld modifier to merge split edges during interpolation, and a vertex group to select the skirt. I have not yet added full handling and logic for detecting edges with big angles like the skirt, or for handling boundaries like on the hair, so both those areas can get better too. You can also see that while it successfully smooths out the Face, it isn't really stylistically correct there. That is still best done with a proxy mesh to define a new shape. This is part of the tools I am working on for Fondant. We are putting together a Blender Addon to release this + a proxy mesh tool for the face, and are working on resolving other problems in-engine to fully bring dynamic light to real time 3D toon shading. Give us a follow, and send them a DM if you are interested in testing these tools as they develop!

aVersionOfReality

14,679 views • 1 year ago

🚨Architects are going to hate this. Someone just open sourced a full 3D building editor that runs entirely in your browser. No AutoCAD. No Revit. No $5,000/year licenses. It's called Pascal Editor. Built with React Three Fiber and WebGPU -- meaning it renders directly on your GPU at near-native speed. Here's what's inside this thing: → A full building/level/wall/zone hierarchy you can edit in real time → An ECS-style architecture where every object updates through GPU-powered systems → Zustand state management with full undo/redo built in → Next.js frontend so it deploys as a web app, not a desktop install → Dirty node tracking -- only re-renders what changed, not the whole scene Here's the wildest part: You can stack, explode, or solo individual building levels. Select a zone, drag a wall, reshape a slab -- all in 3D, all in the browser. Architecture firms pay $50K+ per seat for BIM software that does this workflow. This is free. 100% Open Source.

🚨Architects are going to hate this. Someone just open sourced a full 3D building editor that runs entirely in your browser. No AutoCAD. No Revit. No $5,000/year licenses. It's called Pascal Editor. Built with React Three Fiber and WebGPU -- meaning it renders directly on your GPU at near-native speed. Here's what's inside this thing: → A full building/level/wall/zone hierarchy you can edit in real time → An ECS-style architecture where every object updates through GPU-powered systems → Zustand state management with full undo/redo built in → Next.js frontend so it deploys as a web app, not a desktop install → Dirty node tracking -- only re-renders what changed, not the whole scene Here's the wildest part: You can stack, explode, or solo individual building levels. Select a zone, drag a wall, reshape a slab -- all in 3D, all in the browser. Architecture firms pay $50K+ per seat for BIM software that does this workflow. This is free. 100% Open Source.

Guri Singh

2,967,409 views • 5 months ago

I’m excited to share Deputies, an open source control plane for background coding agents! The goal is simple: move agentic software development off individual laptops and into a central hub where teams can see what agents are doing, review their outputs, and keep a traceable history of the work. Inspired by systems like Inspect/Open-Inspect, Minions, and Open SWE, I built Deputies on top of fks's Flue Framework. It’s designed to run in your own infrastructure and trust boundaries, with Slack/GitHub integrations and Daytona + Docker sandbox support.

I’m excited to share Deputies, an open source control plane for background coding agents! The goal is simple: move agentic software development off individual laptops and into a central hub where teams can see what agents are doing, review their outputs, and keep a traceable history of the work. Inspired by systems like Inspect/Open-Inspect, Minions, and Open SWE, I built Deputies on top of fks's Flue Framework. It’s designed to run in your own infrastructure and trust boundaries, with Slack/GitHub integrations and Daytona + Docker sandbox support.

Sid Palas

12,852 views • 3 months ago

Wow. Recreating the Shawshank Redemption prison in 3D from a single video, in real time (!) Just read the MASt3R-SLAM paper and it's pretty neat. These folks basically built a real-time dense SLAM system on top of MASt3R, which is a transformer-based neural network that can do 3d reconstruction and localization from uncalibrated image pairs. The cool part is they don't need a fixed camera model -- it just works with arbitrary cameras -- think different focal lengths, sensor sizes, even handling zooming in video (FMV drone video anyone?!). If you've done photogrammetry or played with NeRFs you know that is a HUGE deal. They've solved some tricky problems like efficient point matching and tracking, plus they've figured out how to fuse point clouds and handle loop closures in real-time. Their system runs at about 15 FPS on a 4090 and produces both camera poses and dense geometry. When they know the camera calibration, they get SOTA results across several benchmarks, but even without calibration, they still perform well. What's interesting is the approach -- most recent SLAM work has built on DROID-SLAM's architecture, but these folks went a different direction by leveraging a strong 3D reconstruction prior. Seems to give them more coherent geometry, which makes sense since that's what MASt3R was designed for. For anyone who cares about monocular SLAM and 3D reconstruction, this feels like a significant step toward plug-and-play dense SLAM without calibration headaches -- perfect for drones, robots, AR/VR -- the works!

Wow. Recreating the Shawshank Redemption prison in 3D from a single video, in real time (!) Just read the MASt3R-SLAM paper and it's pretty neat. These folks basically built a real-time dense SLAM system on top of MASt3R, which is a transformer-based neural network that can do 3d reconstruction and localization from uncalibrated image pairs. The cool part is they don't need a fixed camera model -- it just works with arbitrary cameras -- think different focal lengths, sensor sizes, even handling zooming in video (FMV drone video anyone?!). If you've done photogrammetry or played with NeRFs you know that is a HUGE deal. They've solved some tricky problems like efficient point matching and tracking, plus they've figured out how to fuse point clouds and handle loop closures in real-time. Their system runs at about 15 FPS on a 4090 and produces both camera poses and dense geometry. When they know the camera calibration, they get SOTA results across several benchmarks, but even without calibration, they still perform well. What's interesting is the approach -- most recent SLAM work has built on DROID-SLAM's architecture, but these folks went a different direction by leveraging a strong 3D reconstruction prior. Seems to give them more coherent geometry, which makes sense since that's what MASt3R was designed for. For anyone who cares about monocular SLAM and 3D reconstruction, this feels like a significant step toward plug-and-play dense SLAM without calibration headaches -- perfect for drones, robots, AR/VR -- the works!

Bilawal Sidhu

704,124 views • 1 year ago

I built a Three.js rendering study inspired by Tiny Glade’s painterly aesthetic, and got it running at 120fps in the browser. Over the past few weeks, I’ve been studying how stylized games achieve that soft, handcrafted look in real time. Tiny Glade was a huge inspiration, and I wanted to use the browser as a constraint: no compute shaders, no native GPU access, and single-threaded JavaScript. As part of this study, I implemented: - GPU-driven instanced brick walls with procedural noise jitter and elastic build animations - Tree, bush, and flower rendering with billboard card expansion, wind sway, and grow animations - Procedural grass with terrain conformance and interactive push deformation - Animated water with layered noise, interactive ripples, and Fresnel-based reflections - Procedural terrain with slope-aware triplanar materials, dirt paths, and rocks - A 7-pass post-processing stack with TAA, bloom, depth of field, painterly filtering, ACES tonemapping, 3D LUT color grading, and film grain The hardest part wasn’t writing any single shader. It was making all of these systems work together at high frame rates inside WebGL, where every millisecond counts and performance problems compound quickly across animation, materials, post-processing, and scene management. Some techniques in this study were inspired by analyzing Tiny Glade’s rendering approach, while others were original implementations built from scratch from visual reference. That contrast taught me a lot: recreating an effect is one challenge, but designing your own shaders and systems to achieve a similar feel is a very different one. This is a private educational rendering study. Some temporary placeholder content is being used during the research phase, and any public or production version would use original or properly licensed assets. Huge credit to Pounce Light for the incredible art direction and rendering work in Tiny Glade: Three.js #gamedev #webgl #threejs #rendering #graphics #realtimerendering #shaderdev

I built a Three.js rendering study inspired by Tiny Glade’s painterly aesthetic, and got it running at 120fps in the browser. Over the past few weeks, I’ve been studying how stylized games achieve that soft, handcrafted look in real time. Tiny Glade was a huge inspiration, and I wanted to use the browser as a constraint: no compute shaders, no native GPU access, and single-threaded JavaScript. As part of this study, I implemented: - GPU-driven instanced brick walls with procedural noise jitter and elastic build animations - Tree, bush, and flower rendering with billboard card expansion, wind sway, and grow animations - Procedural grass with terrain conformance and interactive push deformation - Animated water with layered noise, interactive ripples, and Fresnel-based reflections - Procedural terrain with slope-aware triplanar materials, dirt paths, and rocks - A 7-pass post-processing stack with TAA, bloom, depth of field, painterly filtering, ACES tonemapping, 3D LUT color grading, and film grain The hardest part wasn’t writing any single shader. It was making all of these systems work together at high frame rates inside WebGL, where every millisecond counts and performance problems compound quickly across animation, materials, post-processing, and scene management. Some techniques in this study were inspired by analyzing Tiny Glade’s rendering approach, while others were original implementations built from scratch from visual reference. That contrast taught me a lot: recreating an effect is one challenge, but designing your own shaders and systems to achieve a similar feel is a very different one. This is a private educational rendering study. Some temporary placeholder content is being used during the research phase, and any public or production version would use original or properly licensed assets. Huge credit to Pounce Light for the incredible art direction and rendering work in Tiny Glade: Three.js #gamedev #webgl #threejs #rendering #graphics #realtimerendering #shaderdev

Ibrahim Boona

58,625 views • 4 months ago

$Great question! 🤔 How do you simulate *multiple* layers of glass/refraction in video games? In the last breakdown, I discussed how to create a glass shader in Unity URP. In essence, we were taking the render of the scene from the camera without any transparent objects. This is available in URP as the global _CameraOpaqueTexture. This is good enough for most use-cases, and more or less the classic way of doing it. 🔍 What is _CameraOpaqueTexture? As the name implies, there are no transparent objects rendered into this texture, so it's not possible by default to have something like a transparent-type ocean material/shader rendered through a refractive glass shader (which samples and distorts this texture to render on its surface, as if it's transparent). ⚠️ Why it’s tricky: It's much easier to sort without much further setup if you don't have refraction, and only a transparent material, because in that case you're not simulating the transparency yourself via sampling the rendered scene texture. But for refraction, it's required-- unless you want to go down the ray/path tracing route. You could simulate accurate, real dispersion... and that's about as expensive as it sounds, and it requires a rework of your entire rendering. --> 🚫 It's not a viable suggestion to offer. 📚 There are well-known terms regarding transparency sorting you can search up, but as you've specifically asked for refractive boxes, I'll discuss briefly about that. 🧱 Simulating layers of refraction: For this kind of rendering, you need some way to render the backfaces before rendering the front. And the backfaces that are rendered may contain whatever data you'd like for additional processing in the layer front-facing mesh render. 🧪 Examples: You could render the back face as a glass shader of its own, as an intermediate step after _CameraOpaqueTexture. Then you sample this texture instead and you end up with multi-layered refraction, "just like that". You can also render the back normals only, via a fully opaque shader, and use that to manually account for that during the front render. You could even bake in data needed for thickness in realtime. 🛠️ Without making it complicated for yourself, the most straightforward method is via render textures, and you can easily set some fractional resolution. Cameras in Unity have an open slot for target textures to render to. You can use custom render textures to process _SelfTexture2D. ⏱️ It's great to do low-resolution processing for more complex tasks, like blurring and caustics. You can get massive performance boosts, considering the square law and number of pixels/fragments that need calculations (quadratic scaling). 🚧 I've not fully exploited the possibilities myself, but research/development with PRISM is ongoing!$

Great question! 🤔 How do you simulate multiple layers of glass/refraction in video games? In the last breakdown, I discussed how to create a glass shader in Unity URP. In essence, we were taking the render of the scene from the camera without any transparent objects. This is available in URP as the global _CameraOpaqueTexture. This is good enough for most use-cases, and more or less the classic way of doing it. 🔍 What is _CameraOpaqueTexture? As the name implies, there are no transparent objects rendered into this texture, so it's not possible by default to have something like a transparent-type ocean material/shader rendered through a refractive glass shader (which samples and distorts this texture to render on its surface, as if it's transparent). ⚠️ Why it’s tricky: It's much easier to sort without much further setup if you don't have refraction, and only a transparent material, because in that case you're not simulating the transparency yourself via sampling the rendered scene texture. But for refraction, it's required-- unless you want to go down the ray/path tracing route. You could simulate accurate, real dispersion... and that's about as expensive as it sounds, and it requires a rework of your entire rendering. --> 🚫 It's not a viable suggestion to offer. 📚 There are well-known terms regarding transparency sorting you can search up, but as you've specifically asked for refractive boxes, I'll discuss briefly about that. 🧱 Simulating layers of refraction: For this kind of rendering, you need some way to render the backfaces before rendering the front. And the backfaces that are rendered may contain whatever data you'd like for additional processing in the layer front-facing mesh render. 🧪 Examples: You could render the back face as a glass shader of its own, as an intermediate step after _CameraOpaqueTexture. Then you sample this texture instead and you end up with multi-layered refraction, "just like that". You can also render the back normals only, via a fully opaque shader, and use that to manually account for that during the front render. You could even bake in data needed for thickness in realtime. 🛠️ Without making it complicated for yourself, the most straightforward method is via render textures, and you can easily set some fractional resolution. Cameras in Unity have an open slot for target textures to render to. You can use custom render textures to process _SelfTexture2D. ⏱️ It's great to do low-resolution processing for more complex tasks, like blurring and caustics. You can get massive performance boosts, considering the square law and number of pixels/fragments that need calculations (quadratic scaling). 🚧 I've not fully exploited the possibilities myself, but research/development with PRISM is ongoing!

Mirza Beig

61,468 views • 1 year ago

SOMEONE JUST BUILT AN OPERATING SYSTEM FOR ROBOTS not a framework... not a library an actual OS dimos is an open-source platform that lets you build, run, and deploy software on any robot humanoids, quadrupeds, drones, wheeled bots you write python modules .... they talk to each other via typed streams...just like microservices, but for physical hardware the wild part..? you can give your AI editor (cursor, claude code) MCP access to send real physical commands "move forward 1 meter" → robot moves. it ships with a full navigation stack SLAM, collision avoidance, route planning. plus a dashboard to debug, visualize, and control your robot in real-time via waypoints, VR, or keyboard. supported hardware already includes unitree G1 humanoids, dji drones, and multiple robot arms. programming physical robots is finally as simple as programming software

SOMEONE JUST BUILT AN OPERATING SYSTEM FOR ROBOTS not a framework... not a library an actual OS dimos is an open-source platform that lets you build, run, and deploy software on any robot humanoids, quadrupeds, drones, wheeled bots you write python modules .... they talk to each other via typed streams...just like microservices, but for physical hardware the wild part..? you can give your AI editor (cursor, claude code) MCP access to send real physical commands "move forward 1 meter" → robot moves. it ships with a full navigation stack SLAM, collision avoidance, route planning. plus a dashboard to debug, visualize, and control your robot in real-time via waypoints, VR, or keyboard. supported hardware already includes unitree G1 humanoids, dji drones, and multiple robot arms. programming physical robots is finally as simple as programming software

Vaishnavi

19,745 views • 4 months ago

3D scanning and rendering is moving so fast - got my splats up and running and I'm mind blown getting ~100fps for this complex 3D scene ⬇️ 🤯 1. WAY faster than NeRF: For comparison, NeRFs would takes around 10 seconds per frame (!) Instead I'm zipping around with FPV controls without breaking a sweat - though I do crash a few times towards the end of the video lol 2. Old Meets New: Gaussian Splatting is cool in that it fuses classical graphics and deep learning techniques. Like NeRFs, this is still a radiance field - just without the slower (ne)ural rendering part. 3. Explicit Representation: Instead you represent a 3D scene as a collection of ellipsoidal "splats" called gaussians. Each gaussian has a position, size, and color. Rendering in real-time is done by projecting into the image plane and alpha blending. 4. Photorealistic Effects: Gaussian splatting use spherical harmonics to represent the view-dependent effects and lighting - allowing surfaces to change color when viewed from different angles, enabling greater photorealism. It doesn't use a neural network, but the training loop is similar to deep learning. 5. Enables Direct Editing: But it's not just speed - with Gaussian Splatting you also get 3D editing support! So you can select, move, and delete stuff, even relight stuff. This type of editing has been more tedious to do with NeRFs and their implicit black box representations. 📲 More tests cooking! Much more to unpack here including simpler explanations. If you enjoyed this post, you might enjoy my feed: Bilawal Sidhu

3D scanning and rendering is moving so fast - got my splats up and running and I'm mind blown getting ~100fps for this complex 3D scene ⬇️ 🤯 1. WAY faster than NeRF: For comparison, NeRFs would takes around 10 seconds per frame (!) Instead I'm zipping around with FPV controls without breaking a sweat - though I do crash a few times towards the end of the video lol 2. Old Meets New: Gaussian Splatting is cool in that it fuses classical graphics and deep learning techniques. Like NeRFs, this is still a radiance field - just without the slower (ne)ural rendering part. 3. Explicit Representation: Instead you represent a 3D scene as a collection of ellipsoidal "splats" called gaussians. Each gaussian has a position, size, and color. Rendering in real-time is done by projecting into the image plane and alpha blending. 4. Photorealistic Effects: Gaussian splatting use spherical harmonics to represent the view-dependent effects and lighting - allowing surfaces to change color when viewed from different angles, enabling greater photorealism. It doesn't use a neural network, but the training loop is similar to deep learning. 5. Enables Direct Editing: But it's not just speed - with Gaussian Splatting you also get 3D editing support! So you can select, move, and delete stuff, even relight stuff. This type of editing has been more tedious to do with NeRFs and their implicit black box representations. 📲 More tests cooking! Much more to unpack here including simpler explanations. If you enjoyed this post, you might enjoy my feed: Bilawal Sidhu

Bilawal Sidhu

337,090 views • 2 years ago

George Kurtz: An iPhone is safer than a regular computer — but if a nation state wants to get in bad enough, they’re gonna get in. There’s a whole black and gray market for zero-day vulnerabilities. One unknown flaw in iOS can be worth millions — and sometimes all it takes is opening a browser. There are zero days in all software. If you have one for Apple iOS, that’s super valuable. Nation states don’t need phishing emails or fake links. In some cases, just receiving a message is enough to run code and implant malware. If there is a software vulnerability that hasn’t been found and patched, they can exploit that and run code. No click. No warning. Just an SMS that silently installs an implant on your phone. This is what modern cyber warfare actually looks like People think phones are safe by default — but there were vulnerabilities where they could just send you an SMS. You wouldn’t even know it happened. The message processes, the exploit runs, and suddenly your phone isn’t yours anymore. If they want to get in bad enough, they’re gonna get in. That’s the reality of zero-days, nation states, and why ‘secure’ doesn’t mean untouchable in modern software.

George Kurtz: An iPhone is safer than a regular computer — but if a nation state wants to get in bad enough, they’re gonna get in. There’s a whole black and gray market for zero-day vulnerabilities. One unknown flaw in iOS can be worth millions — and sometimes all it takes is opening a browser. There are zero days in all software. If you have one for Apple iOS, that’s super valuable. Nation states don’t need phishing emails or fake links. In some cases, just receiving a message is enough to run code and implant malware. If there is a software vulnerability that hasn’t been found and patched, they can exploit that and run code. No click. No warning. Just an SMS that silently installs an implant on your phone. This is what modern cyber warfare actually looks like People think phones are safe by default — but there were vulnerabilities where they could just send you an SMS. You wouldn’t even know it happened. The message processes, the exploit runs, and suddenly your phone isn’t yours anymore. If they want to get in bad enough, they’re gonna get in. That’s the reality of zero-days, nation states, and why ‘secure’ doesn’t mean untouchable in modern software.

Ian Miles Cheong

20,065 views • 6 months ago

I gave a talk at GPU MODE workshop last week on llm.c - the origin story of llm.c - being naked in the world without PyTorch and having to re-invent Array, Autograd, Device, Dtype, Compile, Distributed - how to port a PyTorch layer to 1) explicit PyTorch - and then to 2) write the backward pass - 3) port forward & backward pass to C - 4) string all the layers together - achieving one file of C with no dependencies that compiles and runs ~instantly, where all memory is pre-planned and allocated a single time, fully deterministic, portable code that can run on a potato or a von Neumann probe - how most of llm.c was built at 1am-7am in a water villa porch in Maldives and why this is the recommended way to develop software - convert all of it to run in CUDA on GPU in fp32 - port matmul to cuBLAS - port attention to cuDNN flash-attention - introduce bfloat16 mixed precision - introduce many more optimizations and features like kernel fusions, Packed128, stochastic rounding, full determinism - add multi-GPU training, NCCL, sharded optimizer - add multi-node with MPI or file system or socket - reproduce GPT-2 (1.6B) on one 8XH100 node in 24 hours for $672 in llm.c, achieving (at the time) 29% less memory, 19% faster training that PyTorch nightly, and much faster compile & run - how open source development attracts Avengers from the internet - port to training Llama 3 imminent (branch exists) - many other notable forks - last thought: how software abstractions like Python/PyTorch and everything else really exist only because humans are finite in knowledge, IQ and attention, and how with increasing AI capability LLMs may export custom binaries like llm.c for any application directly, tearing apart and refactoring all abstractions as needed. More links in reply

I gave a talk at GPU MODE workshop last week on llm.c - the origin story of llm.c - being naked in the world without PyTorch and having to re-invent Array, Autograd, Device, Dtype, Compile, Distributed - how to port a PyTorch layer to 1) explicit PyTorch - and then to 2) write the backward pass - 3) port forward & backward pass to C - 4) string all the layers together - achieving one file of C with no dependencies that compiles and runs ~instantly, where all memory is pre-planned and allocated a single time, fully deterministic, portable code that can run on a potato or a von Neumann probe - how most of llm.c was built at 1am-7am in a water villa porch in Maldives and why this is the recommended way to develop software - convert all of it to run in CUDA on GPU in fp32 - port matmul to cuBLAS - port attention to cuDNN flash-attention - introduce bfloat16 mixed precision - introduce many more optimizations and features like kernel fusions, Packed128, stochastic rounding, full determinism - add multi-GPU training, NCCL, sharded optimizer - add multi-node with MPI or file system or socket - reproduce GPT-2 (1.6B) on one 8XH100 node in 24 hours for $672 in llm.c, achieving (at the time) 29% less memory, 19% faster training that PyTorch nightly, and much faster compile & run - how open source development attracts Avengers from the internet - port to training Llama 3 imminent (branch exists) - many other notable forks - last thought: how software abstractions like Python/PyTorch and everything else really exist only because humans are finite in knowledge, IQ and attention, and how with increasing AI capability LLMs may export custom binaries like llm.c for any application directly, tearing apart and refactoring all abstractions as needed. More links in reply

Andrej Karpathy

337,263 views • 1 year ago

🎙️ Want a full breakdown on why cost per token is a key metric for AI infrastructure TCO? We’re operating in a power-constrained world. Demand for intelligence is scaling faster than infrastructure can keep up—so metrics like cost per GPU hour or FLOPs per dollar don’t tell the full story. When it comes to TCO, Cost per Token is the metric that captures end-to-end performance across GPUs, CPUs, networking, software, and ecosystem making it the key driver of real profitability and scalability in AI. If you’re not measuring efficiency at the token level, you’re missing what actually drives cost. Watch the deep-dive conversation with Dr. Gerro Prinsloo , Nader Khalil and Carter Abdallah, live on the NVIDIA YouTube channel.

🎙️ Want a full breakdown on why cost per token is a key metric for AI infrastructure TCO? We’re operating in a power-constrained world. Demand for intelligence is scaling faster than infrastructure can keep up—so metrics like cost per GPU hour or FLOPs per dollar don’t tell the full story. When it comes to TCO, Cost per Token is the metric that captures end-to-end performance across GPUs, CPUs, networking, software, and ecosystem making it the key driver of real profitability and scalability in AI. If you’re not measuring efficiency at the token level, you’re missing what actually drives cost. Watch the deep-dive conversation with Dr. Gerro Prinsloo , Nader Khalil and Carter Abdallah, live on the NVIDIA YouTube channel.

NVIDIA

21,970 views • 4 months ago

In the AI era, the next great software distribution company won’t look like an app store For the last decade-plus, building a strong software business was mostly a game of scarcity. The hard part was assembling the capital, taste, engineering, and distribution to build one enduring app. The discovery systems we built matched that world: app stores, SEO, search bars, rankings, reviews, ads, marketplaces AI has blown up these constraints Soon, anyone will be able to generate software for anything: a workout injury, a work project, a personal workflow, a one-off analysis, a weird niche need only they have, maybe only in a single moment Some apps will last forever, but most won’t. Many will exist for minutes or hours, and be created even faster But just because software can be created at the speed of thought doesn’t mean everyone will become a developer. Most people won’t want to prompt, build, debug, deploy, or manage software. They’ll just want to be matched with the right tool for the job in front of them The same technological wave enabling software creation at scale is also breaking software distribution at scale. Search assumes the best answer already exists somewhere. App stores assume software is something you browse, compare, install, and keep In a world of personal software, the real question becomes: Who understands enough about me, or has earned enough trust from me, to know what software should exist for me right now? The next great app distribution companies may look less like search engines and more like trusted relationships Some will win by importing and processing personal context: your goals, constraints, calendar, body, files, workflows, history, collaborators, and intent. Others will win by exporting taste and trust: experts, creators, communities, agents, and brands that people rely on to decide what is worth using Either way, the new gateway won’t be a search box or an app store shelf. It will be the layer that recommends, assembles, routes, and personalizes software at the moment of need This is the new frontier: context, recommendation, trust, and distribution in the age of infinite software If you’re building platforms or experiences for this future, we want to meet you. Apply to a16z speedrun 🧊. Applications for SR007 close this weekend Application link below

In the AI era, the next great software distribution company won’t look like an app store For the last decade-plus, building a strong software business was mostly a game of scarcity. The hard part was assembling the capital, taste, engineering, and distribution to build one enduring app. The discovery systems we built matched that world: app stores, SEO, search bars, rankings, reviews, ads, marketplaces AI has blown up these constraints Soon, anyone will be able to generate software for anything: a workout injury, a work project, a personal workflow, a one-off analysis, a weird niche need only they have, maybe only in a single moment Some apps will last forever, but most won’t. Many will exist for minutes or hours, and be created even faster But just because software can be created at the speed of thought doesn’t mean everyone will become a developer. Most people won’t want to prompt, build, debug, deploy, or manage software. They’ll just want to be matched with the right tool for the job in front of them The same technological wave enabling software creation at scale is also breaking software distribution at scale. Search assumes the best answer already exists somewhere. App stores assume software is something you browse, compare, install, and keep In a world of personal software, the real question becomes: Who understands enough about me, or has earned enough trust from me, to know what software should exist for me right now? The next great app distribution companies may look less like search engines and more like trusted relationships Some will win by importing and processing personal context: your goals, constraints, calendar, body, files, workflows, history, collaborators, and intent. Others will win by exporting taste and trust: experts, creators, communities, agents, and brands that people rely on to decide what is worth using Either way, the new gateway won’t be a search box or an app store shelf. It will be the layer that recommends, assembles, routes, and personalizes software at the moment of need This is the new frontier: context, recommendation, trust, and distribution in the age of infinite software If you’re building platforms or experiences for this future, we want to meet you. Apply to a16z speedrun 🧊. Applications for SR007 close this weekend Application link below

Josh Lu

11,092 views • 3 months ago

🚨 JAY LENO JUST PULLED UP IN A CAR THAT RUNS ON PERFUME OR TEQUILA — AND IT SOUNDS LIKE A GIANT VACUUM Jay Leno just rolled up in one of the strangest cars ever built: the 1963 Chrysler Turbine Car, with bodies hand-built by Ghia in Italy. Instead of a normal piston engine, it uses a jet-style turbine. And when it starts up, it doesn’t rumble like a V8. It sounds like a giant vacuum cleaner spooling up. Then Leno shows the part that catches everyone off guard first. He puts a bottle of water on the engine while it’s running… and it barely ripples. Almost no vibration at all while the turbine spins. Then he explains why this car is so bizarre. The engine is designed to run on almost anything that burns: • Diesel • Kerosene • Jet fuel • Heating oil • Cooking oil Leno says when Chrysler took the car to Mexico, they ran it on tequila. When they brought it to France, they filled the tank with Chanel No.5 perfume. Only 9 of these Chrysler Turbine Cars are believed to still exist today. If your car could run on almost anything that burns... what would you try putting in the tank first?

🚨 JAY LENO JUST PULLED UP IN A CAR THAT RUNS ON PERFUME OR TEQUILA — AND IT SOUNDS LIKE A GIANT VACUUM Jay Leno just rolled up in one of the strangest cars ever built: the 1963 Chrysler Turbine Car, with bodies hand-built by Ghia in Italy. Instead of a normal piston engine, it uses a jet-style turbine. And when it starts up, it doesn’t rumble like a V8. It sounds like a giant vacuum cleaner spooling up. Then Leno shows the part that catches everyone off guard first. He puts a bottle of water on the engine while it’s running… and it barely ripples. Almost no vibration at all while the turbine spins. Then he explains why this car is so bizarre. The engine is designed to run on almost anything that burns: • Diesel • Kerosene • Jet fuel • Heating oil • Cooking oil Leno says when Chrysler took the car to Mexico, they ran it on tequila. When they brought it to France, they filled the tank with Chanel No.5 perfume. Only 9 of these Chrysler Turbine Cars are believed to still exist today. If your car could run on almost anything that burns... what would you try putting in the tank first?

HustleBitch

868,355 views • 5 months ago