Dart: Towards Redundancy-Free RTL Simulation via DAG-Driven Execution
Open Access
- 26 July 2026
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Abstract
Register Transfer Level (RTL) simulation is key to pre-silicon verification, yet its performance has stagnated in the face of rapidly escalating SoC complexity. Although recent CPU- and GPU-based simulators improve throughput through per-stimulus optimization or batch parallelism, they execute each stimulus in isolation and therefore fail to exploit the substantial computation locality that emerges across stimuli. As many stimuli converge to identical internal states, large portions of the circuit are redundantly re-evaluated, forming a principal bottleneck to scalable RTL simulation. This paper introduces Dart, a Directed Acyclic Graph (DAG)-driven RTL simulation framework that systematically eliminates cross-stimulus redundancy. Dart constructs a DAG-based intermediate representation that makes structural commonality and shared subexpressions across stimuli explicit, enabling principled redundancy elimination through systematic sub-DAG merging. A computation-centric execution engine evaluates shared logic once and amortizes its results across all stimuli that traverse the corresponding state, while a lightweight state-reconstruction mechanism preserves per-stimulus correctness with negligible overhead. Across a suite of industrial RTL designs, Dart delivers speedups of up to 136.7× over Verilator and 4.1× over RTLflow, respectively.Keywords
Funding Information
- National Key Research and Development Program of China (2023YFB4503400)
- National Natural Science Foundation of China (62502172)
- National Natural Science Foundation of China (62322205)
- Zhejiang Provincial Natural Science Foundation of China (LY24F020014)
- Fundamental and Interdisciplinary Disciplines Breakthrough Plan of the Ministry of Education of China (JYB2025XDXM103)
This publication has 17 references indexed in Scilit:
- GSIM: Accelerating RTL Simulation for Large-Scale Designs2025
- GEM: GPU-Accelerated Emulator-Inspired RTL Simulation2025
- GenFuzz: GPU-accelerated Hardware Fuzzing using Genetic Algorithm with Multiple Inputs2023
- DPU-v2: Energy-efficient execution of irregular directed acyclic graphs2022
- Efficiently Exploiting Low Activity Factors to Accelerate RTL Simulation2020
- A Case for Accelerating Software RTL SimulationIEEE Micro, 2020
- Yosys+nextpnr: An Open Source Framework from Verilog to Bitstream for Commercial FPGAs2019
- Integrating NVIDIA Deep Learning Accelerator (NVDLA) with RISC-V SoC on FireSim2019
- Pyverilog: A Python-Based Hardware Design Processing Toolkit for Verilog HDL2015
- Cramming More Components Onto Integrated CircuitsProceedings of the IEEE, 1998