catalyst-neuromorphic

catalyst-neuromorphic / catalyst-n1

Public

Catalyst N1 — Open source neuromorphic processor (Loihi 1 parity). 128 cores, 131K neurons, 14-opcode learning ISA, FPGA-validated on AWS F2.

19
3
100% credibility
Found Mar 23, 2026 at 19 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Verilog
AI Summary

Catalyst N1 is an open-source design for a neuromorphic processor that simulates brain-like spiking neural networks using Python software and optional hardware.

How It Works

1
🕵️ Discover Catalyst N1

You stumble upon this free project online that lets anyone build brain-like computer models.

2
📥 Grab the toolkit

Download the simple software kit and set it up on your computer in moments.

3
🧠 Build your brain network

Create groups of virtual neurons, connect them, and set how they learn from experiences.

4
⚡ Watch it come alive

Hit run and see the neurons fire sparks, learning patterns right before your eyes.

5
📊 Test cool examples

Try recognizing voices or hand movements with ready-made brain models.

6
Choose your speed
💻
Computer simulation

Everything runs smoothly on your regular setup.

⚙️
Hardware boost

Load onto a chip for lightning-quick real-world power.

🎉 Your AI brain succeeds

Enjoy efficient, learning networks that handle tough tasks like a real brain.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 19 to 19 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is catalyst-n1?

Catalyst N1 delivers an open-source neuromorphic processor design in Verilog, hitting Loihi 1 parity with up to 128 cores and 131k neurons using LIF models and STDP learning on a 14-opcode ISA. FPGA-validated on AWS F2 instances, it scales from Arty A7 boards to cloud deploys. Python SDK unifies SNN workflows: build networks, sim on CPU/GPU, run benchmarks like SHD digits or DVS gestures, and flash to hardware via UART/PCIe.

Why is it gaining traction?

Unlike proprietary neuromorphic chips, this open catalyst github repo gives full RTL access plus a drop-in SDK for sim-to-chip parity—no vendor lock-in. Developers swap backends seamlessly, hitting 8k+ timesteps/sec on AWS F2 with real ML tasks. Custom 14-opcode learning ISA lets you tweak rules without recompiling silicon.

Who should use this?

Neuromorphic researchers porting Loihi workloads to FPGAs. SNN engineers training SHD classifiers or DVS gesture nets on GPU before AWS deploy. Hardware tinkerers building low-power edge AI on Arty boards.

Verdict

Promising open catalyst github repo at 19 stars and 1.0% credibility—docs shine, benchmarks pass regression, but low adoption signals early maturity. Fire up sdk benchmarks for a quick win if neuromorphic calls.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.