Alternatives
Products that do what I built a lite LPU that can do inference on Karpathy's MicroGPT does
We had no guide or course that teaches chip design at our university. We had taken a digital logic course, but were disappointed with the fact that the most complex project we did was building a full adder in Quartus using logic blocks, not even in RTL!5 Therefore, we decided to challenge ourselves to dive deep into machine learning (ML) hardware and learn as much as we could on our own. We wanted to prove that basic math (like y = mx + b) and basic logic circuits are enough to help anyone understand how modern AI hardware works. Our goal was to design our own version of the LPU from scratch…
- 1
- 2IB
Built a ~9M param LLM from scratch to understand how they actually work. Vanilla transformer, 60K synthetic conversations, ~130 lines of PyTorch. Trains in 5 min on a free Colab T4. The fish thinks the meaning of life is food. Fork it and swap the personality for your own character.
Apr 2026 · github.com
- 3IB
We wanted to do something very challenging to prove to ourselves that we can do anything we put our mind to. The reasoning for why we chose to build a toy TPU specifically is fairly simple: - Building a chip for ML workloads seemed cool - There was no well-documented open source repo for an ML accelerator that performed both inference and training None of us have real professional experience in hardware design, which, in a way, made the TPU even more appealing since we weren't able to estimate exactly how difficult it would be. As we worked on the initial stages of this project, we…
2025 · tinytpu.com
- 4
- 5IM
2022 · imihajlov.tk
- 6AT
A 3.16M-parameter INT4 transformer running entirely in the on-chip memory of a Xilinx Kria KV260. Zero DRAM in the token loop, 59,965 tok/s on the fabric, bit-exact. Chat with it live.
27d ago · mikeayles.com
- 7SU
Here's a project I've been working on for the last few months. It's a new (I think) algorithm, that allows to adjust smoothly - and in real time - how many calculations you'd like to do during inference of an LLM model. It seems that it's possible to do just 20-25% of weight multiplications instead of all of them, and still get good inference results. I implemented it to run on M1/M2/M3 GPU. The mmul approximation itself can be pushed to run 2x fast before the quality of output collapses. The inference speed is just a bit faster than Llama.cpp's, because the rest of implementation…
2024 · asciinema.org
- 8TV
May 2026 · github.com
- 9ZΜ
How small can a language model be while still doing something useful? I wanted to find out, and had some spare time over the holidays. Z80-μLM is a character-level language model with 2-bit quantized weights ({-2,-1,0,+1}) that runs on a Z80 with 64KB RAM. The entire thing: inference, weights, chat UI, it all fits in a 40KB .COM file that you can run in a CP/M emulator and hopefully even real hardware! It won't write your emails, but it can be trained to play a stripped down version of 20 Questions, and is sometimes able to maintain the illusion of having simple but terse conversations…
Dec 2025 · github.com
- 10
- 11D3
I replicated David Ng's RYS method (https://dnhkng.github.io/posts/rys/) on consumer AMD GPUs (RX 7900 XT + RX 6950 XT) and found something I didn't expect. Transformers appear to have discrete "reasoning circuits" — contiguous blocks of 3-4 layers that act as indivisible cognitive units. Duplicate the right block and the model runs its reasoning pipeline twice. No weights change. No training. The model just thinks longer. The results on standard benchmarks (lm-evaluation-harness, n=50): Devstral-24B, layers 12-14 duplicated once: - BBH Logical Deduction: 0.22 → 0.76…
Mar 2026 · github.com
- 12IM
This is a little toy project of mine that lets you simulate digital logic graphs. It was inspired by Minecraft's Redstone and the Piet esolang. It's got some serious drawbacks-- you write circuits as PNGs and simulate them with a Python interface. It's slow to run and slow to experiment with. And it is certainly difficult to use for people with any kind of color blindness. But despite that, I hope this can still be a fun toy!
2022 · github.com
- 13
General Compute▲315AI models that run on an inference cloud optimized for speed
May 2026 · generalcompute.com
- 14TA
2020 · thinc.ai
- 15EL
2023 · github.com
- 16WM
We wrote our inference engine on Rust, it is faster than llama cpp in all of the use cases. Your feedback is very welcomed. Written from scratch with idea that you can add support of any kernel and platform.
2025 · github.com
- 17IM
Inspired by games like Turing Complete/Virtual Circuit Board/Logic World, I tried to make a mix of aseprite and wiredworld/wired-logic, the idea being the user can build a digital circuit using a "fullstack" pixelart creation workflow. The circuit is just an image. The primitives are (i) connected wires which have undefined, 0 or 1 state during simulation (displayed brighter or darker in function of the state) and (ii) NANDs, which are little pixel triangles. During simulation the user can interact with any wire by clicking on it, toggling its state, which is cool for messing…
2024 · github.com
- 18IU
Hi Hacker News, This is definitely out of my comfort zone. I just wanted to show you guys because I'm super proud of it. It's a 100% faithful recreation based off of the schematics, patents, and ROMs that were found online. So please watch the video and tell me what you think https://youtu.be/auOlZXI1VxA The reason why I think this is relevant is because I've been a programmer for 25 years and AI scares the shit out of me. I'm not a programmer anymore. I'm something else now. I don't know what it is but it's multi-disciplinary, and it doesn't involve writing code myself--for…
Jan 2026
- 19IB
I’ve spent the last few months building a deep learning engine completely from scratch in Python (using only math and random). What started as a basic linear algebra calculator project grew into a symbolic tensor system with autodiff, custom matrix ops, attention mechanisms, LayerNorm, GELU, and even a text generation demo trained on the Brown corpus. I'm still an undergrad, so my main goal is to deeply understand how deep learning actually works under the hood - gradients, attention, backpropagation, optimizers - by building it step-by-step with full visibility into everything, and without…
2025 · github.com
- 20BA
I am a Dutch high school student and last year I had to do a large project of my own choice. Together with two other guys I decided to do a project about digital logic. We had an ambitious plan: building an 8-bits computer. At that time I started learning programming and I already knew some HTML and CSS. So we decided to not build a physical computer, but one running in a self-built simulator. I was going to build the simulator, the other guys took care of the computer. I was hard, our school hadn't taught us anything about computers and programming so we had to learn everything ourselves.…
2017
- 21LO
Hi HN, I’m Joe. My friends Matthew, Jake and I are building Luminal (https://luminalai.com/), a GPU compiler for automatically generating fast GPU kernels for AI models. It uses search-based compilation to achieve high performance. We take high level model code, like you'd have in PyTorch, and generate very fast GPU code. We do that without using LLMs or AI - rather, we pose it as a search problem. Our compiler builds a search space, generates millions of possible kernels, and then searches through it to minimize runtime. You can try out a demo in `demos/matmul` on mac to…
2025 · github.com
- 22FD
I worked on this applied Deep Reinforcement Learning course for the better part of 2021. I made a Datacamp course [0] before, and this served as my inspiration to make an applied Deep RL series. Normally, Deep RL courses teach a lot of mathematically involved theory. You get the practical applications near the end (if at all). I have tried to turn that on its head. In the top-down approach, you learn practical skills first, then go deeper later. This is much more fun. This course (the first in a planned multi-part series) shows how to use the Deep Reinforcement Learning framework RLlib to…
2022 · courses.dibya.online
- 23
- 24TR
2018 · actcast.io
Ranked by how close each launch is in meaning, then by votes. Refine with a description →