CODMAS: A Dialectic Multi-Agent Collaborative Framework for Structured RTL Optimization

Che-Ming Chang; Prashanth Vijayaraghavan; Ashutosh Jadhav; Charles Mackin; Vandana Mukherjee; Hsinyu Tsai; Ehsan Degan

cs.CLMar 17, 2026

CODMAS: A Dialectic Multi-Agent Collaborative Framework for Structured RTL Optimization

Authors: Che-Ming Chang, Prashanth Vijayaraghavan, Ashutosh Jadhav, Charles Mackin, Vandana Mukherjee, Hsinyu Tsai, Ehsan Degan

Organizations: National Taiwan University · IBM Research

Abstract

Optimizing Register Transfer Level (RTL) code is a critical step in Electronic Design Automation (EDA) for improving power, performance, and area (PPA). We present CODMAS (Collaborative Optimization via a Dialectic Multi-Agent System), a framework that combines structured dialectic reasoning with domain-aware code generation and deterministic evaluation to automate RTL optimization. At the core of CODMAS are two dialectic agents: the Articulator, inspired by rubber-duck debugging, which articulates stepwise transformation plans and exposes latent assumptions; and the Hypothesis Partner, which predicts outcomes and reconciles deviations between expected and actual behavior to guide targeted refinements. These agents direct a Domain-Specific Coding Agent (DCA) to generate architecture-aware Verilog edits and a Code Evaluation Agent (CEA) to verify syntax, functionality, and PPA metrics. We introduce RTLOPT, a benchmark of 120 Verilog triples (unoptimized, optimized, testbench) for pipelining and clock-gating transformations. Across proprietary and open LLMs, CODMAS achieves ~25% reduction in critical path delay for pipelining and ~22% power reduction for clock gating, while reducing functional and compilation failures compared to strong prompting and agentic baselines. These results demonstrate that structured multi-agent reasoning can significantly enhance automated RTL optimization and scale to more complex designs and broader optimization tasks.

Figures & tables

Figure 1: Overview of the CodMas framework, illustrating dialectic interaction between agents and iterative refinement of RTL code.

RTLOpt Dataset Statistics
# Pipelining Code Triples	70
# Clock Gating Code Triples	50
Power Range $(nW)$	$\sim{1}-\sim{19,000}$
Area Range $(\mu m^{2})$	$\sim{1}-\sim{18,000}$
Delay Range $(ns)$	$\sim{15}-\sim{1,600}$

Table 1: Summary of the RTLOpt Dataset for evaluating pipelining and clock gating optimizations.

Figure 2: Illustration of the complete CodMas architecture, showing dialectic agents (Articulator & Hypothesis Partner), the executor agents (Domain-Specific Coding Agent & Code Evaluation Agent), and the iterative RTL refinement loop (e.g., pipelined array multiplier). The example highlights a pipelined array multiplier optimization, with feedback from different components guiding successive refinements toward improved PPA metrics.

Pipelining
Models	Zero-Shot			Prompting/Agentic			LLM-VeriPPA			CodMas
	A ( $\downarrow$ )	T ( $\uparrow$ )	FR ( $\downarrow$ )	A ( $\downarrow$ )	T ( $\uparrow$ )	FR ( $\downarrow$ )	A ( $\downarrow$ )	T ( $\uparrow$ )	FR ( $\downarrow$ )	A ( $\downarrow$ )	T ( $\uparrow$ )	FR ( $\downarrow$ )
GPT-4o	0.985	11.4	54.2	0.982	15.2	49.6	0.986	15.4	39.7	0.960	25.5	19.5
GPT-3.5-turbo	0.989	10.0	58.0	0.988	14.0	51.2	1.008	13.6	43.3	0.972	21.3	23.4
DeepSeek-v2.5	1.009	9.8	57.8	0.986	12.8	50.0	1.025	13.0	42.8	0.979	21.4	22.8
Llama-3	1.045	8.7	61.4	1.181	11.6	53.8	1.116	11.3	47.1	1.032	19.8	25.5

Table 2: Performance comparison on pipelining and clock gating (CG). Pipelining reports Area (A), Timing (T), and Failure Rate (FR); clock gating reports A, Power (P), and FR. Bold indicates best;

\uparrow

higher is better,

\downarrow

lower is better. All improvements are statistically significant (

p<0.01

) via paired t-tests with standard deviation in Table 6 .

	Full	w/o CEA	w/o DA	w/o DKI
Pipelining
T (GPT-4o)	25.5	17.2	12.9	11.5
FR (GPT-4o)	19.5	31.0	38.5	32.3
T (Llama-3)	19.8	12.6	10.4	9.3
FR (Llama-3)	25.5	36.9	44.7	35.8
Clock Gating

Table 3: Component-wise ablation study of CodMas . DA: Dialectic Agents; DKI: Domain Knowledge Injection; CEA: Code Evaluation Agent.

Figure 3: Ablation comparing CODMAS with a shared-memory multi-agent (SMA) variant where the Articulator and Hypothesis Partner roles are combined.

Figure 4: Impact of iterative refinement in CodMas on pipelining timing improvement (%) for easy (left) and hard (right) RTLOpt problems. Early iterations yield substantial gains; later iterations show diminishing or unstable returns, especially for harder cases.

Appendix figures & tables3 assets

Supplementary material from the paper’s appendix.

Appendix

Dataset	Size	Functionality	Synthesizability	Metric-specific
VerilogEval Liu et al. (2023)	156	✓	✗	✗
RTLLM Lu et al. (2024)	30	✓	✓	✗
RTLRewriter Yao et al. (2024)	95	✗	✓	✗
TuRTLe Garcia-Gasulla et al. (2025)	223	✓	✓	✗
CVDP Pinckney et al. (2025)	783	✓	✓	✗
ResBench Guo and Zhao (2025)	56	✓	✓	✗

Appendix

Table 4: Comparison of RTLOpt with prior Verilog datasets. “Metric-specific” indicates whether metric-specific optimization exists or not for evaluation.

Module	Original RTL	CodMas Optimized RTL
Multiplier	Sequential multiply w/o pipeline	Partial product pipelined with inter-stage registers
Adder	Ripple-carry 32-bit adder	Pipelined adder with reduced critical path
Control Logic	Unconditioned enable signals	Handshake-aware control signals with proper gating
Observed Errors in Baselines	Syntax errors, functional mismatches, unmet PPA targets
Corrected by CodMas	All syntax errors fixed, functional simulation passes, critical path reduced $\sim 25\%$

Appendix

Table 5: Example of pipelining transformation and error mitigation under CodMas . Dialectic agents guide structured edits, and the CEA validates correctness and PPA improvements.

CodMas Performance: Mean $\pm$ Std over 5 runs
Model	Pipelining			Clock Gating
	A ( $\downarrow$ )	T ( $\uparrow$ )	FR ( $\downarrow$ )	A ( $\downarrow$ )	P ( $\uparrow$ )	FR ( $\downarrow$ )
GPT-4o	0.960 $\pm$ 0.007	25.5 $\pm$ 1.2	19.5 $\pm$ 2.1	0.999 $\pm$ 0.006	21.7 $\pm$ 1.5	21.8 $\pm$ 2.3
GPT-3.5-turbo	0.972 $\pm$ 0.009	21.3 $\pm$ 1.1	23.4 $\pm$ 2.4	1.020 $\pm$ 0.007	18.8 $\pm$ 1.2	24.2 $\pm$ 2.0
DeepSeek-v2.5	0.979 $\pm$ 0.010	21.4 $\pm$ 1.0	22.8 $\pm$ 2.3	1.015 $\pm$ 0.008	19.0 $\pm$ 1.3	23.6 $\pm$ 2.1
Llama-3	1.032 $\pm$ 0.012	19.8 $\pm$ 1.3	25.5 $\pm$ 2.5	1.048 $\pm$ 0.010	16.5 $\pm$ 1.1	26.3 $\pm$ 2.4

Appendix

Table 6: Extracted CodMas results with mean and standard deviation over five runs. Metrics: area (A), timing (T), power (P), and failure rate (FR). FR denotes fraction of runs failing syntax, functional, or PPA checks.

Electronic Design Automation

Electronic Design Automation uses algorithms and software to design, verify, optimize, and manufacture electronic circuits and systems.

See all 214 papers

Dr. RTL: Autonomous Agentic RTL Optimization through Tool-Grounded Self-Improvement
Apr 16, 2026Wenji Fang, Yao Lu, Shang Liu +5Large Language Model-Guided Optimization Electronic Design Automation
Hong Kong University of Science and Technology (HKUST)
VeriAgent: A Tool-Integrated Multi-Agent System with Evolving Memory for PPA-Aware RTL Code Generation
Mar 18, 2026Yaoxiang Wang, Qi Shi, ShangZhan Li +6RTL Code Generation Electronic Design Automation
Xiamen University · Tsinghua University · Harbin Institute of Technology
ChipCraftBrain: Validation-First RTL Generation via Multi-Agent Orchestration
Apr 21, 2026Cagri EryilmazElectronic Design Automation Code Generation
ChipCraftX San Francisco, CA

Apr 16, 2026J/K move · Enter open · S save

Dr. RTL: Autonomous Agentic RTL Optimization through Tool-Grounded Self-Improvement

Wenji Fang, Yao Lu, Shang Liu, Jing Wang+4

Hong Kong University of Science and Technology (HKUST)

Large Language Model-Guided Optimization Electronic Design Automation LLM Agent Skill Learning

Open paper AskPDF arXiv

Recent advances in large language models (LLMs) have sparked growing interest in automatic RTL optimization for better performance, power, and area (PPA). However, existing methods are still far from realistic RTL optimization. Their evaluation settings are often unrealistic: they are tested on manually degraded, small-scale RTL designs and rely on weak open-source tools. Their optimization methods are also limited, relying on coarse design-level feedback and simple pre-defined rewriting rules. To address these limitations, we present Dr. RTL, an agentic framework for RTL timing optimization in a realistic evaluation environment, with continual self-improvement through reusable optimization skills. We establish a realistic evaluation setting with more challenging RTL designs and an industrial EDA workflow. Within this setting, Dr. RTL performs closed-loop optimization through a multi-agent framework for critical-path analysis, parallel RTL rewriting, and tool-based evaluation. We further introduce group-relative skill learning, which compares parallel RTL rewrites and distills the optimization experience into an interpretable skill library. Currently, this library contains 47 pattern--strategy entries for cross-design reuse to improve PPA and accelerate convergence, and it can continue evolving over time. Evaluated on 20 real-world RTL designs, Dr. RTL achieves average WNS/TNS improvements of 21%/17% with a 6% area reduction over the industry-leading commercial synthesis tool.

Abstract

Electronic Design Automation

Explore similar work

Dr. RTL: Autonomous Agentic RTL Optimization through Tool-Grounded Self-Improvement

VeriAgent: A Tool-Integrated Multi-Agent System with Evolving Memory for PPA-Aware RTL Code Generation

ChipCraftBrain: Validation-First RTL Generation via Multi-Agent Orchestration