Search papers, labs, and topics across Lattice.
RefEvo, a multi-agent framework, tackles challenges in LLM-based hardware reference model generation by dynamically adapting workflows, co-evolving verification logic, and compressing context. It uses a Dynamic Design Planner to decompose specifications, a Co-Evolutionary Verification Mechanism with a Dialectical Arbiter to rectify models and testbenches, and a Spec Anchoring Strategy for lossless context compression. Experiments on 20 hardware modules demonstrate a 95% pass rate, significantly outperforming static baselines, and achieves 71.04% token reduction with complete specification recall.
LLMs can now generate reliable hardware reference models with 95% accuracy thanks to a novel co-evolutionary verification mechanism that weeds out correlated hallucinations between model and testbench.
As the complexity of System-on-Chip (SoC) designs grows, the shift-left paradigm necessitates the rapid development of high-fidelity reference models (typically written in SystemC) for early architecture exploration and verification. While Large Language Models (LLMs) show promise in code generation, their application to hardware modeling faces unique challenges: (1) Rigid, static workflows fail to adapt to varying design complexity, causing inefficiency; (2) Context window overflow in multi-turn interactions leads to catastrophic forgetting of critical specifications; and (3) the Coupled Validation Failure problem--where generated Testbenches (TBs) incorrectly validate flawed models due to correlated hallucinations--severely undermines reliability. To address these limitations, we introduce RefEvo, a dynamic multi-agent framework designed for agile and reliable reference modeling. RefEvo features three key innovations: (1) A Dynamic Design Planner that autonomously decomposes design specifications and constructs tailored execution workflows based on semantic complexity; (2) A Co-Evolutionary Verification Mechanism, which employs a Dialectical Arbiter to simultaneously rectify the model and verification logic against the specification (Spec) oracle, effectively mitigating false positives; and (3) A Spec Anchoring Strategy for lossless context compression. Evaluated on a diverse benchmark of 20 hardware modules, RefEvo achieves a 95% pass rate, outperforming static baselines by a large margin. Furthermore, our context optimization reduces token consumption by an average of 71.04%, achieving absolute savings of over 70,000 tokens per session for complex designs while maintaining 100% specification recall.