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.

The best way to learn CUDA is by solving problems. In this video, I implement 7 beginner CUDA kernels from scratch: • Vector Addition • Matrix Transpose • Matrix Addition • RGB to Grayscale • Value Clipping • Sigmoid Activation • Interleave Arrays If you're getting started with CUDA... show more

voxium

1,133 subscribers

36,987 views • 17 days ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Introducing The AI CUDA Engineer: An agentic AI system that automates the production of highly optimized CUDA kernels. The AI CUDA Engineer can produce highly optimized CUDA kernels, reaching 10-100x speedup over common machine learning operations in PyTorch. Our system is also able to produce highly optimized CUDA kernels that are much faster than existing CUDA kernels commonly used in production. We believe that fundamentally, AI systems can and should be as resource-efficient as the human brain, and that the best path to achieve this efficiency is to use AI to make AI more efficient! We are excited to publish our paper, The AI CUDA Engineer: Agentic CUDA Kernel Discovery, Optimization and Composition. We also release a dataset of over 17,000 verified CUDA kernels produced by The AI CUDA Engineer. Paper: Kernel Archive Webpage: HuggingFace Dataset: The AI CUDA Engineer utilizes evolutionary LLM-driven code optimization to autonomously improve the runtime of machine learning operations. Our system is not only able to convert PyTorch code into CUDA kernels, but through the use of evolution, it can also optimize the runtime performance of CUDA kernels, fuse multiple operations, and even discover novel solutions for writing efficient CUDA operations by learning from past innovations! We believe The AI CUDA Engineer opens a new era of AI-driven acceleration of AI and automated inference time optimization. We (Robert Lange, Aaditya Prasad 🇺🇸, sssss, Maxence Faldor, Yujin Tang, hardmaru) are excited to continue Sakana AI's mission of leveraging AI to improve AI.

Introducing The AI CUDA Engineer: An agentic AI system that automates the production of highly optimized CUDA kernels. The AI CUDA Engineer can produce highly optimized CUDA kernels, reaching 10-100x speedup over common machine learning operations in PyTorch. Our system is also able to produce highly optimized CUDA kernels that are much faster than existing CUDA kernels commonly used in production. We believe that fundamentally, AI systems can and should be as resource-efficient as the human brain, and that the best path to achieve this efficiency is to use AI to make AI more efficient! We are excited to publish our paper, The AI CUDA Engineer: Agentic CUDA Kernel Discovery, Optimization and Composition. We also release a dataset of over 17,000 verified CUDA kernels produced by The AI CUDA Engineer. Paper: Kernel Archive Webpage: HuggingFace Dataset: The AI CUDA Engineer utilizes evolutionary LLM-driven code optimization to autonomously improve the runtime of machine learning operations. Our system is not only able to convert PyTorch code into CUDA kernels, but through the use of evolution, it can also optimize the runtime performance of CUDA kernels, fuse multiple operations, and even discover novel solutions for writing efficient CUDA operations by learning from past innovations! We believe The AI CUDA Engineer opens a new era of AI-driven acceleration of AI and automated inference time optimization. We (Robert Lange, Aaditya Prasad 🇺🇸, sssss, Maxence Faldor, Yujin Tang, hardmaru) are excited to continue Sakana AI's mission of leveraging AI to improve AI.

Sakana AI

1,160,190 views • 1 year ago

Day 14/90 of Inference Engineering I didn't finish the MNIST in CUDA project today, which made me hesitant to post since I didn't want to put out some slop. But in the spirit of being consistent and showing up, here's a concept I've revisited today now that I've started writing CUDA again. This is an animation on the CUDA memory life cycle. Generally speaking, when writing CUDA, memory management goes through the following stages: 1. Allocate memory on the host 2.Initialize / populate the memory 3.Allocate device memory 4.Copy host data to the device 5. Launch the kernel 6. Copy device data back to the host 7. Free the memory - On a brief reflection, I'm starting to write the matmul kernels in CUDA instead of in C. So this in itself is another paradigm shift in thinking. Excited to hone my technique over the coming week. I'll update you guys when I've completed the project; otherwise I'll just make a post like usual on what I've learned during the day!

Day 14/90 of Inference Engineering I didn't finish the MNIST in CUDA project today, which made me hesitant to post since I didn't want to put out some slop. But in the spirit of being consistent and showing up, here's a concept I've revisited today now that I've started writing CUDA again. This is an animation on the CUDA memory life cycle. Generally speaking, when writing CUDA, memory management goes through the following stages: 1. Allocate memory on the host 2.Initialize / populate the memory 3.Allocate device memory 4.Copy host data to the device 5. Launch the kernel 6. Copy device data back to the host 7. Free the memory - On a brief reflection, I'm starting to write the matmul kernels in CUDA instead of in C. So this in itself is another paradigm shift in thinking. Excited to hone my technique over the coming week. I'll update you guys when I've completed the project; otherwise I'll just make a post like usual on what I've learned during the day!

max fu

42,808 views • 17 days ago

Day 9/90 of Inference Engineering I built a visualization tool for CUDA kernels and C code to view the buffers for each step through. Check out Yesterday, I was terramogged by a simple implementation of a basic matmul between three matrices in 1D vector form. There were essentially three 1D vectors, each vector representing a corresponding matrix (A, B, C). In the implementation, each matrix is just just a pointer to the first element of a 1D array. You are to compute the output matrix through A^T @ B and store it in C. This morning I built a visualizer to help me see and understand why the triple for-loops work the way they do. In the coming days, I’m confident I can make sense as well as backpropagation! As for it does not include a built-in compiler. It's just an interpreter; the visualization shows exactly what the code does, bugs included. The main thing I wanted to visualize is matmul with a triple for-loop between a sequential program to that of a parallel one. Expect bugs and other inconsistencies within I’ll polish this mini-project in the coming days. As for the rest of today, I’ll be trying to understand backpropagation in C and CUDA .

Day 9/90 of Inference Engineering I built a visualization tool for CUDA kernels and C code to view the buffers for each step through. Check out Yesterday, I was terramogged by a simple implementation of a basic matmul between three matrices in 1D vector form. There were essentially three 1D vectors, each vector representing a corresponding matrix (A, B, C). In the implementation, each matrix is just just a pointer to the first element of a 1D array. You are to compute the output matrix through A^T @ B and store it in C. This morning I built a visualizer to help me see and understand why the triple for-loops work the way they do. In the coming days, I’m confident I can make sense as well as backpropagation! As for it does not include a built-in compiler. It's just an interpreter; the visualization shows exactly what the code does, bugs included. The main thing I wanted to visualize is matmul with a triple for-loop between a sequential program to that of a parallel one. Expect bugs and other inconsistencies within I’ll polish this mini-project in the coming days. As for the rest of today, I’ll be trying to understand backpropagation in C and CUDA .

max fu

12,564 views • 22 days ago

The latest MLX has a CUDA back-end! To get started: pip install "mlx[cuda]" With the same codebase you can develop locally, run your model on Apple silicon, or in the cloud on Nvidia GPUs. MLX is designed around Apple silicon - which has a unified memory architecture. It uses the same design with CUDA. So there's no need to move arrays around from CPU memory to GPU memory. Note, this is early days - some operations are missing and performance is still being optimized. But it's already quite fast for Transformer training, text generation, and more! Here's a demo using mlx-lm to generate text with Llama 3 8B (bf16) on an A100:

The latest MLX has a CUDA back-end! To get started: pip install "mlx[cuda]" With the same codebase you can develop locally, run your model on Apple silicon, or in the cloud on Nvidia GPUs. MLX is designed around Apple silicon - which has a unified memory architecture. It uses the same design with CUDA. So there's no need to move arrays around from CPU memory to GPU memory. Note, this is early days - some operations are missing and performance is still being optimized. But it's already quite fast for Transformer training, text generation, and more! Here's a demo using mlx-lm to generate text with Llama 3 8B (bf16) on an A100:

Awni Hannun

42,761 views • 1 year ago

I trained a 12M parameter LLM on my own ML framework using a Rust backend and CUDA kernels for flash attention, AdamW, and more. Wrote the full transformer architecture, and BPE tokenizer from scratch. The framework features: - Custom CUDA kernels (Flash Attention, fused LayerNorm, fused GELU) for 3x increased throughput - Automatic WebGPU fallback for non-NVIDIA devices - TypeScript API with Rust compute backend - One npm install to get started, prebuilt binaries for every platform Try out the model for yourself: Built with Reese Chong. Check out the repos and blog if you want to learn more. Shoutout to Modal for the compute credits allowing me to train on 2 A100 GPUs without going broke cc sunny madra Gavin

I trained a 12M parameter LLM on my own ML framework using a Rust backend and CUDA kernels for flash attention, AdamW, and more. Wrote the full transformer architecture, and BPE tokenizer from scratch. The framework features: - Custom CUDA kernels (Flash Attention, fused LayerNorm, fused GELU) for 3x increased throughput - Automatic WebGPU fallback for non-NVIDIA devices - TypeScript API with Rust compute backend - One npm install to get started, prebuilt binaries for every platform Try out the model for yourself: Built with Reese Chong. Check out the repos and blog if you want to learn more. Shoutout to Modal for the compute credits allowing me to train on 2 A100 GPUs without going broke cc sunny madra Gavin

Aadi Kulshrestha

814,352 views • 3 months ago

you can't tell me AI isn't a game changer for programming it's connecting over webrtc the screen is made of 16 by 16 regions each region is 256 by 256 ive come up with a custom encoding I spam UDP packets for each reigion for each client each region is designed to fit in a single packet the next state of the simulation is computed in my GPU using cuda the cuda is linked to golang the program manages a double buffer this is so that the network broadcastloop doesn't contend with the cuda update for the object when you write it sends a udp packet for every mouse event i manage a stream to update it on my gpu the client takes each packet and has custom webgl to update the regions and render everything in a single pass. texture 3d

you can't tell me AI isn't a game changer for programming it's connecting over webrtc the screen is made of 16 by 16 regions each region is 256 by 256 ive come up with a custom encoding I spam UDP packets for each reigion for each client each region is designed to fit in a single packet the next state of the simulation is computed in my GPU using cuda the cuda is linked to golang the program manages a double buffer this is so that the network broadcastloop doesn't contend with the cuda update for the object when you write it sends a udp packet for every mouse event i manage a stream to update it on my gpu the client takes each packet and has custom webgl to update the regions and render everything in a single pass. texture 3d

kache

97,819 views • 1 year ago

I no longer recommend the lunge matrix and leg swings as the best way to warm up for a run 🤯 Here's the best dynamic warm-up that I know of from Jeff Boele Coaches: This is the time to learn and share it with your athletes.

I no longer recommend the lunge matrix and leg swings as the best way to warm up for a run 🤯 Here's the best dynamic warm-up that I know of from Jeff Boele Coaches: This is the time to learn and share it with your athletes.

Jay Johnson

102,553 views • 2 years ago

Cerebras just had the biggest IPO of the year. Founder Andrew Feldman says the 3 most important things he had to convince investors of while doing the roadshow were that demand for inference is going to 1,000,000x, the GPU isn't the only way to do compute, and that the CUDA moat is overstated. What he said: "Jensen said some time ago on Brad Gerstner's podcast that the demand for inference will grow by a 1,000,000x, and nobody believed him. And at the same time, you saw Sam Altman displaying real vision and going out and trying to lock up huge amounts of compute, memory, data centers, and power, because he saw it too." "[We tried] to share what that means — what exponential demand means. And that we're still so early, and yet the demand for AI compute is overwhelming." "The other thing is that there are lots of ways to do this. The GPU isn't the only way. You've got TPUs, Trainium, and us. There are lots of different ways to build a solution here." "And finally — the notion that CUDA is this grand lock-in is overplayed. Gemini 3, which is an excellent model, was trained on TPUs with no CUDA. The Anthropic models were trained on Trainium with no CUDA. Some of the best models, some of the most interesting things are being done without CUDA. And that lock-in might be overplayed." $CBRS

Cerebras just had the biggest IPO of the year. Founder Andrew Feldman says the 3 most important things he had to convince investors of while doing the roadshow were that demand for inference is going to 1,000,000x, the GPU isn't the only way to do compute, and that the CUDA moat is overstated. What he said: "Jensen said some time ago on Brad Gerstner's podcast that the demand for inference will grow by a 1,000,000x, and nobody believed him. And at the same time, you saw Sam Altman displaying real vision and going out and trying to lock up huge amounts of compute, memory, data centers, and power, because he saw it too." "[We tried] to share what that means — what exponential demand means. And that we're still so early, and yet the demand for AI compute is overwhelming." "The other thing is that there are lots of ways to do this. The GPU isn't the only way. You've got TPUs, Trainium, and us. There are lots of different ways to build a solution here." "And finally — the notion that CUDA is this grand lock-in is overplayed. Gemini 3, which is an excellent model, was trained on TPUs with no CUDA. The Anthropic models were trained on Trainium with no CUDA. Some of the best models, some of the most interesting things are being done without CUDA. And that lock-in might be overplayed." $CBRS

TBPN

36,474 views • 2 months ago

Did you know you're not actually "moving" your crypto from an exchange to a cold wallet? This is one of the most common beginner misunderstandings-and in this video, I explain what's really happening in a simple, easy-to-understand way so you can use a cold wallet with confidence

Did you know you're not actually "moving" your crypto from an exchange to a cold wallet? This is one of the most common beginner misunderstandings-and in this video, I explain what's really happening in a simple, easy-to-understand way so you can use a cold wallet with confidence

Sheryl Crypto Advice

40,089 views • 2 months ago

If you're serious about mastering Claude, then this is the most important video you'll ever watch. How to go from Beginner → Pro in 3 steps, with the exact memory system I use to set up Cowork + Claude Code for the best results. This video will permanently change how you use Claude:

If you're serious about mastering Claude, then this is the most important video you'll ever watch. How to go from Beginner → Pro in 3 steps, with the exact memory system I use to set up Cowork + Claude Code for the best results. This video will permanently change how you use Claude:

Miles Deutscher

137,939 views • 4 months ago

Windsurf CEO explains why traditional startup moats are mostly silly. Varun said even with 50-100 of the best MIT engineers, that's still just "hundreds of engineering years" in your product. Someone else can build something similar in the same space. He used Nvidia as an example. People think CUDA is their real moat, but that's not accurate. Large companies like Google, OpenAI, and Anthropic spend tens of billions on chips. If CUDA didn't exist, they'd find a way to write assembly code and make GPUs work. People use CUDA because Nvidia made it really capable at low floating-point math, great interconnect, and fast computers. Not because they're locked in. Every year Nvidia has pressure - if they don't make hardware, interconnect, and memory bandwidth much faster, their profit margins shrink and AMD will compete. For startups, the only moat is speed. Learning where the dead bodies are. Understanding what doesn't work and building compounding advantages through that knowledge. Speed lets you learn from the market faster, which means you're first to the next idea too.

Windsurf CEO explains why traditional startup moats are mostly silly. Varun said even with 50-100 of the best MIT engineers, that's still just "hundreds of engineering years" in your product. Someone else can build something similar in the same space. He used Nvidia as an example. People think CUDA is their real moat, but that's not accurate. Large companies like Google, OpenAI, and Anthropic spend tens of billions on chips. If CUDA didn't exist, they'd find a way to write assembly code and make GPUs work. People use CUDA because Nvidia made it really capable at low floating-point math, great interconnect, and fast computers. Not because they're locked in. Every year Nvidia has pressure - if they don't make hardware, interconnect, and memory bandwidth much faster, their profit margins shrink and AMD will compete. For startups, the only moat is speed. Learning where the dead bodies are. Understanding what doesn't work and building compounding advantages through that knowledge. Speed lets you learn from the market faster, which means you're first to the next idea too.

Aish

334,732 views • 1 year ago

I've migrated the old Mast3r-SLAM example I had made last year to the latest version of Rerun and made a bunch of improvements! I wanted to spend some time with agents to modernize it. Here's an example of me walking around with my iPhone and getting a dense reconstruction at about 10FPS on a 5090. Heres the following improvements I made. Brought it into the monorepo with proper packaging: • Using prefix.dev pixi-build to get rid of all the mast3r/asmk/lietorch vendored code with just a few small patches. This let me remove so 60k lines of code from the repo! • Don't have to build the lietorch code on my machine anymore, which was taking ~10 minutes to compile (and also made it work on blackwell when it previously did not) Rebuilt the Gradio interface: • Fixed incremental updates, .MOV uploads, and stop behavior • Made the CLI + Gradio interface share the same entry point so updates automatically propagate Upgraded the Rerun integration: • Switched to a multiprocessing async logging strategy • Added video/pointmap/confidence logging • Improved blueprint layout and hid noisy entities from 3D view • Biggest perf win was the async background logger - documented about a ~2.5x speedup from decoupling logging from tracking The newest and most interesting part was my attempt to replace the CUDA kernels for Gauss-Newton ray matching with a Modular Mojo backend. As a Python dev, every time I look at CUDA code I basically shy away as it's pretty difficult for me to understand. Mojo let me rewrite the matching logic in a syntax I'm more comfortable with while still getting near-CUDA performance. Mojo is now the default matching backend with CUDA fallback. One major piece that's missing is the custom PyTorch op path, but I'll eventually do that as well. I heavily leaned on Claude Code to do the CUDA → Mojo migration, and I have no doubt it's not the cleanest or most idiomatic, BUT it's way more readable for me and helps me better understand the underlying algorithm. This was a ton of work, and a large part of why I'm doing it is how the monorepo compounds. This becomes an artifact for the next example I want to build with Claude that I can point to, which will make it even faster to implement. The compounding nature of this is really interesting and part of why I'm spending so much time trying to make things nice and readable.

I've migrated the old Mast3r-SLAM example I had made last year to the latest version of Rerun and made a bunch of improvements! I wanted to spend some time with agents to modernize it. Here's an example of me walking around with my iPhone and getting a dense reconstruction at about 10FPS on a 5090. Heres the following improvements I made. Brought it into the monorepo with proper packaging: • Using prefix.dev pixi-build to get rid of all the mast3r/asmk/lietorch vendored code with just a few small patches. This let me remove so 60k lines of code from the repo! • Don't have to build the lietorch code on my machine anymore, which was taking ~10 minutes to compile (and also made it work on blackwell when it previously did not) Rebuilt the Gradio interface: • Fixed incremental updates, .MOV uploads, and stop behavior • Made the CLI + Gradio interface share the same entry point so updates automatically propagate Upgraded the Rerun integration: • Switched to a multiprocessing async logging strategy • Added video/pointmap/confidence logging • Improved blueprint layout and hid noisy entities from 3D view • Biggest perf win was the async background logger - documented about a ~2.5x speedup from decoupling logging from tracking The newest and most interesting part was my attempt to replace the CUDA kernels for Gauss-Newton ray matching with a Modular Mojo backend. As a Python dev, every time I look at CUDA code I basically shy away as it's pretty difficult for me to understand. Mojo let me rewrite the matching logic in a syntax I'm more comfortable with while still getting near-CUDA performance. Mojo is now the default matching backend with CUDA fallback. One major piece that's missing is the custom PyTorch op path, but I'll eventually do that as well. I heavily leaned on Claude Code to do the CUDA → Mojo migration, and I have no doubt it's not the cleanest or most idiomatic, BUT it's way more readable for me and helps me better understand the underlying algorithm. This was a ton of work, and a large part of why I'm doing it is how the monorepo compounds. This becomes an artifact for the next example I want to build with Claude that I can point to, which will make it even faster to implement. The compounding nature of this is really interesting and part of why I'm spending so much time trying to make things nice and readable.

Pablo Vela

42,143 views • 3 months ago

Chamath: US AI Startups Can Learn A Lot from DeepSeek "This is a case where necessity was the mother of invention." On E213, Chamath Palihapitiya explained how, even if the $6M number is inaccurate, DeepSeek still had some impressive breakthroughs: GRPO > PPO for reinforcement learning "These guys were like, 'Well, how am I going to do this reinforcement learning thing?' They invented a totally different algorithm." "It uses a lot less computer memory and it's highly performant." PTX > CUDA to build the model "And then the second thing that was crazy is everybody is used to building models and compiling through CUDA, which is NVIDIA's proprietary language." "And these guys worked totally around CUDA and they did something called PTX, which goes right to the bare metal." What US AI startups can learn from this: "We, meaning the West, with all the money that we've had, didn't come up with these ideas." "And I think part of why we didn't come up is not that we're not smart enough to do it, but we weren't forced to because the constraints didn't exist." "And so I just wonder how we make sure we learn this principle, meaning, when the AI company wakes up and rolls out of bed and some VC gives them $200M, maybe that's not the right answer for a Series A or a Seed." "And maybe the right answer is $2M so that they do these DeepSeek-like innovations."

Chamath: US AI Startups Can Learn A Lot from DeepSeek "This is a case where necessity was the mother of invention." On E213, Chamath Palihapitiya explained how, even if the $6M number is inaccurate, DeepSeek still had some impressive breakthroughs: GRPO > PPO for reinforcement learning "These guys were like, 'Well, how am I going to do this reinforcement learning thing?' They invented a totally different algorithm." "It uses a lot less computer memory and it's highly performant." PTX > CUDA to build the model "And then the second thing that was crazy is everybody is used to building models and compiling through CUDA, which is NVIDIA's proprietary language." "And these guys worked totally around CUDA and they did something called PTX, which goes right to the bare metal." What US AI startups can learn from this: "We, meaning the West, with all the money that we've had, didn't come up with these ideas." "And I think part of why we didn't come up is not that we're not smart enough to do it, but we weren't forced to because the constraints didn't exist." "And so I just wonder how we make sure we learn this principle, meaning, when the AI company wakes up and rolls out of bed and some VC gives them $200M, maybe that's not the right answer for a Series A or a Seed." "And maybe the right answer is $2M so that they do these DeepSeek-like innovations."

The All-In Podcast

108,815 views • 1 year ago

Discrete Fourier Transform by hand ✍️ ~ 12 steps walkthrough below Here is a little-known secret about the DFT and the inverse DFT: it is just matrix multiplication in both directions, one the transpose of the other, exactly like the forward pass and backpropagation I drew in other examples. Goal: recover which cosine waves a signal is made of, using nothing but multiplication and addition. = 1. Given = Three signals written as sums of cosines, and a fourth, X, that we do not know yet. = 2. Frequency matrix F = Let us write the coefficients as a matrix. Each signal is a row, each frequency a column, so A = cos(w) + 2cos(2w) becomes [1, 2, 0, 0]. = 3. Sample the waves = We read the four cosine waves at ten discrete time points. That word "discrete" is the whole difference between this and the continuous transform. = 4. Cosine matrix W = Let us write those samples as a matrix: each frequency a row, each time point a column. = 5. Frequency to time = We multiply F by W. That combines the four cosine waves in the proportions F specifies, and the result T is the three signals as they would look in time. = 6. Transpose = Let us stand each signal up as a column. = 7. Time to frequency = We multiply W by that transpose. Every cell is the dot product of one signal with one cosine wave, which measures how much of that wave the signal contains. Zero means none of it. = 8. Scale = Let us multiply by 2/n, with n = 10. The projections come out five times too large, and this is the correction. = 9. Transpose back = We turn it back around, and it is F again, exactly. That is the check: the transform recovered the coefficients we started from. = 10. Now solve for X = Let us run the same multiplication on the one signal whose recipe we never knew. = 11. Scale = We divide by 5 again. = 12. Transpose back = And X reads [0, 0, 3, 2], which says X = 3cos(3w) + 2cos(4w). Note: I originally drew this to show that the DFT is a special case of a convolution layer, its filters fixed to sine and cosine waves rather than learned. No wonder, then, that a convolution layer free to learn its own filters can be trained to process signals. 💾 Save this post!

Discrete Fourier Transform by hand ✍️ ~ 12 steps walkthrough below Here is a little-known secret about the DFT and the inverse DFT: it is just matrix multiplication in both directions, one the transpose of the other, exactly like the forward pass and backpropagation I drew in other examples. Goal: recover which cosine waves a signal is made of, using nothing but multiplication and addition. = 1. Given = Three signals written as sums of cosines, and a fourth, X, that we do not know yet. = 2. Frequency matrix F = Let us write the coefficients as a matrix. Each signal is a row, each frequency a column, so A = cos(w) + 2cos(2w) becomes [1, 2, 0, 0]. = 3. Sample the waves = We read the four cosine waves at ten discrete time points. That word "discrete" is the whole difference between this and the continuous transform. = 4. Cosine matrix W = Let us write those samples as a matrix: each frequency a row, each time point a column. = 5. Frequency to time = We multiply F by W. That combines the four cosine waves in the proportions F specifies, and the result T is the three signals as they would look in time. = 6. Transpose = Let us stand each signal up as a column. = 7. Time to frequency = We multiply W by that transpose. Every cell is the dot product of one signal with one cosine wave, which measures how much of that wave the signal contains. Zero means none of it. = 8. Scale = Let us multiply by 2/n, with n = 10. The projections come out five times too large, and this is the correction. = 9. Transpose back = We turn it back around, and it is F again, exactly. That is the check: the transform recovered the coefficients we started from. = 10. Now solve for X = Let us run the same multiplication on the one signal whose recipe we never knew. = 11. Scale = We divide by 5 again. = 12. Transpose back = And X reads [0, 0, 3, 2], which says X = 3cos(3w) + 2cos(4w). Note: I originally drew this to show that the DFT is a special case of a convolution layer, its filters fixed to sine and cosine waves rather than learned. No wonder, then, that a convolution layer free to learn its own filters can be trained to process signals. 💾 Save this post!

Tom Yeh

24,442 views • 6 days ago

I created a web app for teaching computer vision basics, allowing users to upload an image or video stream and instantly view it as a numerical pixel matrix, with options to display values in RGB, Hex, or Grayscale formats. It features predefined convolution kernels applied in real-time: - Sobel operator for edge detection (computes gradients to highlight boundaries) - Gaussian Blur: averages neighboring pixels with weights to reduce noise - Sharpen: boosts local contrast to enhance details This setup clearly visualizes how images are represented as matrices or tensors, how convolution slides kernels across pixels, and how operations like gradient computation, blurring, or sharpening modify the data step-by-step. I plan to use demos like this in my upcoming computer vision course to build intuition for fundamental image processing concepts, OpenCV filters, and CNN building blocks (e.g., why kernels extract features).

I created a web app for teaching computer vision basics, allowing users to upload an image or video stream and instantly view it as a numerical pixel matrix, with options to display values in RGB, Hex, or Grayscale formats. It features predefined convolution kernels applied in real-time: - Sobel operator for edge detection (computes gradients to highlight boundaries) - Gaussian Blur: averages neighboring pixels with weights to reduce noise - Sharpen: boosts local contrast to enhance details This setup clearly visualizes how images are represented as matrices or tensors, how convolution slides kernels across pixels, and how operations like gradient computation, blurring, or sharpening modify the data step-by-step. I plan to use demos like this in my upcoming computer vision course to build intuition for fundamental image processing concepts, OpenCV filters, and CNN building blocks (e.g., why kernels extract features).

Marco Franzon

270,367 views • 7 months ago

introducing matrix. the ultimate multi-agent runtime that turns your product into actual revenue and automates 99% of what it takes to run a company. probably the most different, most future-feeling agent product you've ever seen. the harness, the interface, everything is rebuilt from scratch. think of it as the complement to codex and claude code. they write the code. matrix does literally everything else you actually need done. plug in your codex subscription and run it for free, or byok. fully free either way. this video was edited by matrix and voiced by an agent. more real cases coming soon. matrix is in beta preview right now. full launch next week. retweet + comment and we'll send invite codes to the first 200.

introducing matrix. the ultimate multi-agent runtime that turns your product into actual revenue and automates 99% of what it takes to run a company. probably the most different, most future-feeling agent product you've ever seen. the harness, the interface, everything is rebuilt from scratch. think of it as the complement to codex and claude code. they write the code. matrix does literally everything else you actually need done. plug in your codex subscription and run it for free, or byok. fully free either way. this video was edited by matrix and voiced by an agent. more real cases coming soon. matrix is in beta preview right now. full launch next week. retweet + comment and we'll send invite codes to the first 200.

Derek Nee

15,261 views • 1 month ago

I'm starting my own rust tutorial series. This is a series for anyone who looked at some rust code and felt scared by the syntax. This very first episode will get you started with some basics. But the goal is to gradually build up and go to the complex topics. If you're a complete beginner, then definitely give my series a try. Full video link below 👇

I'm starting my own rust tutorial series. This is a series for anyone who looked at some rust code and felt scared by the syntax. This very first episode will get you started with some basics. But the goal is to gradually build up and go to the complex topics. If you're a complete beginner, then definitely give my series a try. Full video link below 👇

meowy🦀

28,085 views • 3 months ago

📝 AI/ML for Biology & Healthcare: A Learning Path This blog is highly inspired by my experience and exploration in the field of ML for biology and healthcare. Everything that I've been studying for the past years and a more structured way to learn the foundational knowledge to tackle problems in this impactful field. Here's what I covered in the blog: — Programming & Engineering — AI & Machine Learning — Mathematics for ML — Healthcare — Biology & Biolomedicine This is what I hoped to have when I first started learning all this. 🔗 link:

📝 AI/ML for Biology & Healthcare: A Learning Path This blog is highly inspired by my experience and exploration in the field of ML for biology and healthcare. Everything that I've been studying for the past years and a more structured way to learn the foundational knowledge to tackle problems in this impactful field. Here's what I covered in the blog: — Programming & Engineering — AI & Machine Learning — Mathematics for ML — Healthcare — Biology & Biolomedicine This is what I hoped to have when I first started learning all this. 🔗 link:

TK • 木下

26,889 views • 8 months ago

In the movie, The Matrix, Agent Smith has the ability to enter any body, at any time, and that often happens when the Matrix is threatened by Neo. This scenario is a reflection of our real world. For example, you're having a conversation with a friend/family member/partner and everything is fine until you mention something like 9/11 was inside job. Now watch what happens, up pops the Agent Smith to replace the person you were just having a conversation with to attack you because all truths threaten the Matrix, which gives them feeling of safety, security and purpose in this "hologram" reality. Most people have had the Agent Smith programing since birth. These programed people become the guardians of the Matrix, protectors of the narrative and enforcers of indoctrination which holds together our false concept of reality like super glue. Think critically, I love you all, X

In the movie, The Matrix, Agent Smith has the ability to enter any body, at any time, and that often happens when the Matrix is threatened by Neo. This scenario is a reflection of our real world. For example, you're having a conversation with a friend/family member/partner and everything is fine until you mention something like 9/11 was inside job. Now watch what happens, up pops the Agent Smith to replace the person you were just having a conversation with to attack you because all truths threaten the Matrix, which gives them feeling of safety, security and purpose in this "hologram" reality. Most people have had the Agent Smith programing since birth. These programed people become the guardians of the Matrix, protectors of the narrative and enforcers of indoctrination which holds together our false concept of reality like super glue. Think critically, I love you all, X

MAVERICK X

392,586 views • 7 months ago

Congrats to the Kimi.ai team! This is awesome. Great to see this level of research coming from open-source frontier model labs. I liked the paper so much I built a Rust implementation of it ;) Full AttnRes + Block AttnRes with two-phase inference, built using Burn (tensor library and Deep Learning Framework, in Rust, by Tracel AI). Runs on CPU, CUDA, Metal, wgpu. Includes an interactive TUI that trains a model live and visualizes depth attention evolving from uniform to selective in real time. Repo link and more on what is implemented in the comments.

Congrats to the Kimi.ai team! This is awesome. Great to see this level of research coming from open-source frontier model labs. I liked the paper so much I built a Rust implementation of it ;) Full AttnRes + Block AttnRes with two-phase inference, built using Burn (tensor library and Deep Learning Framework, in Rust, by Tracel AI). Runs on CPU, CUDA, Metal, wgpu. Includes an interactive TUI that trains a model live and visualizes depth attention evolving from uniform to selective in real time. Repo link and more on what is implemented in the comments.

abdel

95,237 views • 4 months ago