English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

ChipMATE: Two Small AI Agents Co-Write and Cross-Check RTL, Beating a 1600B-Parameter Model

Forum topic · 小凯 · 2026-05-18

Summary

A forum post discusses ChipMATE, a multi-agent reinforcement learning framework for RTL (Verilog) code generation aimed at closing the gap between academic AI benchmarks and real chip-design workflows. ChipMATE pairs two agents: one writes Verilog RTL, the other independently writes a Python reference model, and the two cross-validate each other's outputs instead of relying on an external golden testbench. Training proceeds in two RL-driven stages — independent training to saturation, then joint cooperative training — plus a backtracking mechanism at inference that regenerates from the last decision point when validation fails. Reportedly, ChipMATE reaches 75.0% pass@1 on VerilogEval V2 with a 4B base model and 80.1% with a 9B model, with the 9B version surpassing the 1600B-parameter DeepSeek V4 general model. The post also raises open questions: quality control of the 64.4K self-generated training samples, risk of cross-validation degenerating into mutual confirmation of shared errors, and potential collapse modes in multi-agent RL.

Chip design has an open secret: AI models that generate RTL code perform well in academic benchmarks but are almost never used in industry. The reason is not that the models write bad code — it is a fundamental mismatch between real industrial workflows and academic evaluation setups.

Academic evaluation assumes a perfect golden testbench exists before the model writes code. In reality, testbenches are often developed alongside the design; there is no golden reference until the code is done. Academic evaluation assumes the model can call closed-source APIs. In reality, chip vendors' servers are physically air-gapped — no outbound data allowed. Academic evaluation assumes models cannot be fine-tuned on proprietary codebases. In reality, every chip company holds decades of proprietary RTL, but that data can neither be uploaded to the cloud nor used for training.

ChipMATE, proposed by Yu, Lin, and their team, changes this paradigm at its root. The core design uses two agents: one writes Verilog (the hardware description language), and one writes a Python reference model (a software behavioral description). After the Verilog agent produces an RTL segment, the Python agent independently writes a corresponding reference model. The two outputs are compared against each other — not against an external golden reference; each acts as the other's validation standard. This mirrors a common industrial practice: two engineers independently design the same module, then cross-verify.

Even more critical is the training pipeline, which has two stages. Stage one trains each agent independently, pushing each one's code generation ability to saturation. Stage two is joint training, teaching the pair to collaborate — one agent learns the other's needs, writing code that is easy for the partner to verify, and producing reference models that are easy for the partner to understand. Both stages are driven by reinforcement learning.

At inference time there is also a backtracking mechanism: if one agent's output causes the other agent's validation to fail, the workflow does not continue; instead, it backtracks to the previous decision point and regenerates. This prevents errors from propagating across turns.

The results are striking. ChipMATE achieves 75.0% pass@1 on VerilogEval V2 with a 4B base model, and 80.1% with a 9B model. Every record previously held by self-trained models was broken. More notably, the 9B version of ChipMATE even surpasses DeepSeek V4 — a general-purpose giant with 1600B parameters. Two specialized agents totaling under 20B parameters beat a 1600B general-purpose colossus.

Unclear points: the paper's training data consists of 64.4K samples generated by the framework itself — how is their quality guaranteed? If the reference-model agent has itself learned wrong patterns, cross-validation could become cross-confirmation. Also, the training stability of multi-agent reinforcement learning has been contested in similar systems — could collapse modes emerge, where the two agents settle into a low-quality mutual默契-style equilibrium?

---

References

1. Yu, Z., Lin, Y., Zhou, C., et al. (2026). *ChipMATE: Multi-Agent Training via Reinforcement Learning for Enhanced RTL Generation*. arXiv:2605.12857 [cs.MA]. 2. Thakur, S., et al. (2024). *VerilogEval: Evaluating Large Language Models for Verilog Code Generation*. DAC. 3. Wu, M., et al. (2025). *RTLCoder: Fully Open-Source and Efficient LLM for RTL Code Generation*. DATE.

Tags

#chipmate#rtl-generation#verilog#multi-agent#reinforcement-learning#llm#eda#verilogeval

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/177620298