Research

Paper

AI LLM February 19, 2026

SimulatorCoder: DNN Accelerator Simulator Code Generation and Optimization via Large Language Models

Authors

Yuhuan Xia, Tun Li, Hongji Zhou, Xianfa Zhou, Chong Chen, Ruiyu Zhang

Abstract

This paper presents SimulatorCoder, an agent powered by large language models (LLMs), designed to generate and optimize deep neural network (DNN) accelerator simulators based on natural language descriptions. By integrating domain-specific prompt engineering including In-Context Learning (ICL), Chain-of-Thought (CoT) reasoning, and a multi-round feedback-verification flow, SimulatorCoder systematically transforms high-level functional requirements into efficient, executable, and architecture-aligned simulator code. Experiments based on the customized SCALE-Sim benchmark demonstrate that structured prompting and feedback mechanisms substantially improve both code generation accuracy and simulator performance. The resulting simulators not only maintain cycle-level fidelity with less than 1% error compared to manually implemented counterparts, but also consistently achieve lower simulation runtimes, highlighting the effectiveness of LLM-based methods in accelerating simulator development. Our code is available at https://github.com/xiayuhuan/SimulatorCoder.

Metadata

arXiv ID: 2602.17169
Provider: ARXIV
Primary Category: cs.AR
Published: 2026-02-19
Fetched: 2026-02-21 18:51

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2602.17169v1</id>\n    <title>SimulatorCoder: DNN Accelerator Simulator Code Generation and Optimization via Large Language Models</title>\n    <updated>2026-02-19T08:34:18Z</updated>\n    <link href='https://arxiv.org/abs/2602.17169v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2602.17169v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>This paper presents SimulatorCoder, an agent powered by large language models (LLMs), designed to generate and optimize deep neural network (DNN) accelerator simulators based on natural language descriptions. By integrating domain-specific prompt engineering including In-Context Learning (ICL), Chain-of-Thought (CoT) reasoning, and a multi-round feedback-verification flow, SimulatorCoder systematically transforms high-level functional requirements into efficient, executable, and architecture-aligned simulator code. Experiments based on the customized SCALE-Sim benchmark demonstrate that structured prompting and feedback mechanisms substantially improve both code generation accuracy and simulator performance. The resulting simulators not only maintain cycle-level fidelity with less than 1% error compared to manually implemented counterparts, but also consistently achieve lower simulation runtimes, highlighting the effectiveness of LLM-based methods in accelerating simulator development. Our code is available at https://github.com/xiayuhuan/SimulatorCoder.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.AR'/>\n    <published>2026-02-19T08:34:18Z</published>\n    <arxiv:primary_category term='cs.AR'/>\n    <author>\n      <name>Yuhuan Xia</name>\n    </author>\n    <author>\n      <name>Tun Li</name>\n    </author>\n    <author>\n      <name>Hongji Zhou</name>\n    </author>\n    <author>\n      <name>Xianfa Zhou</name>\n    </author>\n    <author>\n      <name>Chong Chen</name>\n    </author>\n    <author>\n      <name>Ruiyu Zhang</name>\n    </author>\n  </entry>"
}