ScaleRTL: Scaling LLMs with Reasoning Data and Test-Time Compute for Accurate RTL Code Generation

TL;DR

ScaleRTL is a reasoning large language model designed for RTL code generation, leveraging extensive reasoning data and test-time scaling to significantly improve performance over existing models.

Contribution

The paper introduces ScaleRTL, the first reasoning LLM for RTL coding that scales reasoning data and test-time compute, enabling deeper RTL reasoning and better code generation.

Findings

Achieves state-of-the-art results on VerilogEval and RTLLM benchmarks.

Outperforms 18 baselines by up to 18.4% on VerilogEval.

Utilizes a large dataset of 3.5B tokens with chain-of-thought reasoning traces.

Abstract

Recent advances in large language models (LLMs) have enabled near-human performance on software coding benchmarks, but their effectiveness in RTL code generation remains limited due to the scarcity of high-quality training data. While prior efforts have fine-tuned LLMs for RTL tasks, they do not fundamentally overcome the data bottleneck and lack support for test-time scaling due to their non-reasoning nature. In this work, we introduce ScaleRTL, the first reasoning LLM for RTL coding that scales up both high-quality reasoning data and test-time compute. Specifically, we curate a diverse set of long chain-of-thought reasoning traces averaging 56K tokens each, resulting in a dataset of 3.5B tokens that captures rich RTL knowledge. Fine-tuning a general-purpose reasoning model on this corpus yields ScaleRTL that is capable of deep RTL reasoning. Subsequently, we further enhance the performance of ScaleRTL through a novel test-time scaling strategy that extends the reasoning process via iteratively reflecting on and self-correcting previous reasoning steps. Experimental results show that ScaleRTL achieves state-of-the-art performance on VerilogEval and RTLLM, outperforming 18 competitive baselines by up to 18.4% on VerilogEval and 12.7% on RTLLM.