Research

Paper

AI LLM March 05, 2026

NL2GDS: LLM-aided interface for Open Source Chip Design

Authors

Max Eland, Jeyan Thiyagalingam, Dinesh Pamunuwa, Roshan Weerasekera

Abstract

The growing complexity of hardware design and the widening gap between high-level specifications and register-transfer level (RTL) implementation hinder rapid prototyping and system design. We introduce NL2GDS (Natural Language to Layout), a novel framework that leverages large language models (LLMs) to translate natural language hardware descriptions into synthesizable RTL and complete GDSII layouts via the open-source OpenLane ASIC flow. NL2GDS employs a modular pipeline that captures informal design intent, generates HDL using multiple LLM engines and verifies them, and orchestrates automated synthesis and layout. Evaluations on ISCAS'85 and ISCAS'89 benchmark designs demonstrate up to 36% area reduction, 35% delay reduction, and 70% power savings compared to baseline designs, highlighting its potential to democratize ASIC design and accelerate hardware innovation.

Metadata

arXiv ID: 2603.05489
Provider: ARXIV
Primary Category: cs.AR
Published: 2026-03-05
Fetched: 2026-03-06 14:20

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.05489v1</id>\n    <title>NL2GDS: LLM-aided interface for Open Source Chip Design</title>\n    <updated>2026-03-05T18:55:54Z</updated>\n    <link href='https://arxiv.org/abs/2603.05489v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.05489v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>The growing complexity of hardware design and the widening gap between high-level specifications and register-transfer level (RTL) implementation hinder rapid prototyping and system design. We introduce NL2GDS (Natural Language to Layout), a novel framework that leverages large language models (LLMs) to translate natural language hardware descriptions into synthesizable RTL and complete GDSII layouts via the open-source OpenLane ASIC flow. NL2GDS employs a modular pipeline that captures informal design intent, generates HDL using multiple LLM engines and verifies them, and orchestrates automated synthesis and layout. Evaluations on ISCAS'85 and ISCAS'89 benchmark designs demonstrate up to 36% area reduction, 35% delay reduction, and 70% power savings compared to baseline designs, highlighting its potential to democratize ASIC design and accelerate hardware innovation.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.AR'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.CY'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.LO'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='eess.SY'/>\n    <published>2026-03-05T18:55:54Z</published>\n    <arxiv:comment>10 pages, 6 figures</arxiv:comment>\n    <arxiv:primary_category term='cs.AR'/>\n    <author>\n      <name>Max Eland</name>\n    </author>\n    <author>\n      <name>Jeyan Thiyagalingam</name>\n    </author>\n    <author>\n      <name>Dinesh Pamunuwa</name>\n    </author>\n    <author>\n      <name>Roshan Weerasekera</name>\n    </author>\n  </entry>"
}