HN Showcase
Back to browse
GitHub Repository

Universal Stochastic Computing Framework for Neuromorphic Hardware — Rust SIMD engine, Python simulation, Verilog RTL, HDC/VSA, SCPN integration

8 starsPython

SC-NeuroCore – Rust neuromorphic compiler, 512× speedup

by anulum·Feb 12, 2026·2 points·0 comments
Visit ProjectView on HN

AI Analysis

●●SolidWizardryNiche Gem
The Take

They ship a Rust engine plus Python models that claim cycle-exact, bit-true equivalence with Verilog and a verified co-simulation suite — and back it with a 512× speedup and sub-10µs inference numbers. This is a rare, technically ambitious toolkit for building and proving neuromorphic FPGA/CPU stacks, though the AGPL license and narrow domain mean it’s primarily valuable to hardware teams and researchers, not casual ML users.

Category
Hardware
Target Audience

Neuromorphic researchers, FPGA/embedded hardware engineers, systems/ML researchers working on stochastic computing and low‑latency inference

Post Description

SC-NeuroCore is a verified Rust-based neuromorphic compiler that translates Python SNN definitions into optimized, bit-true stochastic bitstream logic for CPU/FPGA.

Key highlights: - 512.4× real-time speedup on LIF neuron updates (vs. legacy Python) - Bit-true equivalence with FPGA hardware (verified co-simulation, 8/8 tests) - Polymorphic engine: HDC/VSA (AVX-512), Petri Nets, fault-tolerant logic - Sub-10 µs inference latency, 40%+ bit-flip resilience - Install: pip install sc-neurocore-engine - Quick start: see notebooks/01_hdc_symbolic_query.ipynb (HDC demo)

GitHub: https://github.com/anulum/sc-neurocore Rust API docs: https://anulum.github.io/sc-neurocore

Built to bridge Python simulation and hardware deployment for neuromorphic computing. Happy to answer questions about the compiler, benchmarks, or verification.

Similar Projects

Open Source●●Solid

SCPN Fusion Core – Tokamak plasma SIM and neuromorphic SNN control

This repo compiles stochastic Petri-net control policies into sub-millisecond spiking LIF networks and pairs them with reduced-order plasma simulators and Rust acceleration — not something you see every day. It ships validation against real equilibria/shot databases and a Streamlit dashboard, so the project feels like a serious research-to-prototype pipeline rather than paper-only ideas; the tradeoff is deliberate reduced-order physics (not a TRANSP/GENE replacement), which is fair for real-time control work.

WizardryNiche GemBold Bet
anulum
204mo ago