Research Paper · 2025

From Natural Language
to Hardware

An automated pipeline transforming natural-language ideas into circuit schematics through LLM-guided C code generation, recursive compilation, and hardware synthesis.

75%
Compilation Success
2.4×
Improvement vs Single-Pass
7
Target Architectures
93%
FPGA Success Rate

Six-Stage Pipeline

From natural language to hardware in six composable, independently replaceable stages.

01

Code Generation

LLM synthesizes C code from natural-language idea with hardware-oriented characteristics.

02

Recursive Validation

gcc compilation with error classification and LLM-driven correction loop (up to 5 iterations).

03

Copyright Injection

License metadata insertion for circuit IP provenance tracking.

04

Test Generation

20 static heuristic tests per source line across 5 categories (UX, regression, unit, blackbox, A-B).

05

Pattern Conformance

10 software design patterns checked including singleton, observer, modularity, and memory safety.

06

Hardware Compilation

Flux API circuit synthesis with graceful degradation to mock JSON generation.

Benchmark Results

Evaluated on 15 benchmark ideas across 5 hardware target architectures.

TargetSuccess RateMean IterationsMean LOC
ASIC80%3.247.8
FPGA93%2.452.3
TPU67%4.138.5
QPU60%4.729.2
OPU75%3.535.6
Overall75%3.640.7

Architecture Choices

Key design decisions and their rationale.

C as Intermediate Representation

C is the most represented systems language in LLM training data. Mature HLS toolchain (Vivado, Intel HLS) provides a proven compilation path to hardware.

Iterative over Single-Pass

Single-pass compilation achieves 31% success rate. Our recursive loop with R=5 achieves 75% — a 2.4× improvement.

Static over Dynamic Testing

Generated code targets hardware unavailable at test time. Static heuristic tests provide coverage without requiring hardware availability.

CLI-First Design

zsh wrapper for developer familiarity and composability. TypeScript backend for type safety across the six-stage pipeline.

The Paper

From Natural Language to Hardware: The Idea2Circuit Pipeline for Automated Circuit Synthesis via LLM-Guided Code Generation and Iterative Compilation

Research Paper · 7 pages

Abstract

We present Idea2Circuit, an automated pipeline that transforms natural-language descriptions of computational ideas into hardware circuit schematics targeting ASIC, FPGA, TPU, QPU, OPU, LPU, and GPU architectures. The system employs a multi-stage architecture comprising LLM-based C code generation, recursive validation and error-corrected compilation using gcc, static heuristic test generation at 20 tests per source line, design pattern conformance checking, and hardware synthesis via the Flux compilation API. The pipeline achieves correct-by-construction C code within a mean of 2.4 validation iterations on a corpus of 15 benchmark ideas, with a 93% success rate for FPGA-targeted circuits.

Shyamal Suhana Chandra·2025

Download & Cite

Download the full paper or copy the BibTeX citation.

Download PDF
Idea2Circuit Paper · 182 KB