Nan Guan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Nan Guan has a total of 382 co-authors. Here are the top co-authors by number of shared publications:
Wang Yi
Uppsala Univ, Dept Informat Technol, S-75236 Uppsala, Sweden
48 shared publicationsQingxu Deng
30 shared publicationsGe Yu
Northeastern University, China
21 shared publicationsMingsong Lv
Department of Computing, The Hong Kong Polytechnic University, Hung Hom, Hong Kong
21 shared publications- Wang Yi
Uppsala University
18 shared publications - Xu Jiang
University of Electronic Science and Technology of China
17 shared publications
Latest Publications
- 1 December 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 75  (12) , 4220-4234
RT-DPU: Scheduling Real-Time Tasks With Heterogeneous Multi-DPU Engines
- 1 December 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 75  (12) , 4220-4234
RT-DPU: Scheduling Real-Time Tasks With Heterogeneous Multi-DPU Engines
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 75  (10) , 1-15
Achieving Predictable Timer Management Under Varying Workload Conditions in ROS 2
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 75  (10) , 1-15
Achieving Predictable Timer Management Under Varying Workload Conditions in ROS 2
- Open Access
- 27 September 2026
- Published by Association for Computing Machinery (ACM)
- p. 99-104
User Space Buffer Cache with the Bells and Whistles of a Kernel Page Cache
- 27 September 2026
- Published by Association for Computing Machinery (ACM)
- p. 99-104
Open AccessUser Space Buffer Cache with the Bells and Whistles of a Kernel Page Cache
- Open Access
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
CircuitDiff: Bridging Netlist Knowledge with RTL based on Graph Denoising Diffusion
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessCircuitDiff: Bridging Netlist Knowledge with RTL based on Graph Denoising Diffusion
- Open Access
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Insights from Verification: Training a Verilog Generation LLM using Reinforcement Learning with Testbench Feedback
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessInsights from Verification: Training a Verilog Generation LLM using Reinforcement Learning with Testbench Feedback
- Open Access
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
ParetoPilot: Global Optimization Reasoning on HLS Design Space Exploration with LLMs
The combinatorial explosion of directives in high-level synthesis (HLS) creates an expansive design space, impeding the agility and productivity of HLS. Previous efforts to construct design space exploration (DSE) methods with heuristics struggle to precisely identify Pareto-optimal designs under tight search budgets. The advanced reasoning capabilities of large language models (LLMs) present a promising avenue for multi-objective optimization. Therefore, we introduce ParetoPilot, an automated DSE framework, which leverages a dedicated LLM to adopt global optimization reasoning to navigate exploration. Guided by optimization strategies, ParetoPilot integrates directive scheduling and quality awareness to rapidly shape broad and concave Pareto fronts. Compared to a SOTA heuristic-based DSE method, ParetoPilot improves the DSE performance by 49.5% and achieves a 2.2× efficiency gain, outperforming the auxiliary LLM DeepSeek-R1 by 81.4%. The Repo. of code: https://github.com/GEAR-SEU/ParetoPilot-DAC-26.
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessParetoPilot: Global Optimization Reasoning on HLS Design Space Exploration with LLMs
The combinatorial explosion of directives in high-level synthesis (HLS) creates an expansive design space, impeding the agility and productivity of HLS. Previous efforts to construct design space exploration (DSE) methods with heuristics struggle to precisely identify Pareto-optimal designs under tight search budgets. The advanced reasoning capabilities of large language models (LLMs) present a promising avenue for multi-objective optimization. Therefore, we introduce ParetoPilot, an automated DSE framework, which leverages a dedicated LLM to adopt global optimization reasoning to navigate exploration. Guided by optimization strategies, ParetoPilot integrates directive scheduling and quality awareness to rapidly shape broad and concave Pareto fronts. Compared to a SOTA heuristic-based DSE method, ParetoPilot improves the DSE performance by 49.5% and achieves a 2.2× efficiency gain, outperforming the auxiliary LLM DeepSeek-R1 by 81.4%. The Repo. of code: https://github.com/GEAR-SEU/ParetoPilot-DAC-26.
- Open Access
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
On-Demand Multi-Task Sparsity for Efficient Large-Model Deployment on Edge Devices
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessOn-Demand Multi-Task Sparsity for Efficient Large-Model Deployment on Edge Devices
- Open Access
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
ChatSVA: Bridging SVA Generation for Hardware Verification via Task-Specific LLMs
Functional verification consumes over 50% of the IC development lifecycle, where SystemVerilog Assertions (SVAs) are indispensable for formal property verification and enhanced simulation-based debugging. However, manual SVA authoring is labor-intensive and error-prone. While Large Language Models (LLMs) show promise, their direct deployment is hindered by low functional accuracy and a severe scarcity of domain-specific data. To address these challenges, we introduce ChatSVA, an end-to-end SVA generation system built upon a multi-agent framework. At its core, the Agent-Bridge platform enables this multi-agent approach by systematically generating high-purity datasets, overcoming the data scarcity inherent to few-shot scenarios. Evaluated on 24 RTL designs, ChatSVA achieves 98.66% syntax and 96.12% functional pass rates, generating 139.5 SVAs per design with 82.50% function coverage. This represents a 33.3 percentage point improvement in functional correctness and an over 11× enhancement in function coverage compared to the previous state-of-the-art (SOTA). ChatSVA not only sets a new SOTA in automated SVA generation but also establishes a robust framework for solving long-chain reasoning problems in few-shot, domain-specific scenarios. A publicly accessible ChatSVA web service is available at https://www.nctieda.com/CHATDV.html.
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessChatSVA: Bridging SVA Generation for Hardware Verification via Task-Specific LLMs
Functional verification consumes over 50% of the IC development lifecycle, where SystemVerilog Assertions (SVAs) are indispensable for formal property verification and enhanced simulation-based debugging. However, manual SVA authoring is labor-intensive and error-prone. While Large Language Models (LLMs) show promise, their direct deployment is hindered by low functional accuracy and a severe scarcity of domain-specific data. To address these challenges, we introduce ChatSVA, an end-to-end SVA generation system built upon a multi-agent framework. At its core, the Agent-Bridge platform enables this multi-agent approach by systematically generating high-purity datasets, overcoming the data scarcity inherent to few-shot scenarios. Evaluated on 24 RTL designs, ChatSVA achieves 98.66% syntax and 96.12% functional pass rates, generating 139.5 SVAs per design with 82.50% function coverage. This represents a 33.3 percentage point improvement in functional correctness and an over 11× enhancement in function coverage compared to the previous state-of-the-art (SOTA). ChatSVA not only sets a new SOTA in automated SVA generation but also establishes a robust framework for solving long-chain reasoning problems in few-shot, domain-specific scenarios. A publicly accessible ChatSVA web service is available at https://www.nctieda.com/CHATDV.html.
- Open Access
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
ZK-Tracer: A High-Performance Heterogeneous Accelerator for Zero-Knowledge VM Trace Generation
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessZK-Tracer: A High-Performance Heterogeneous Accelerator for Zero-Knowledge VM Trace Generation
- Open Access
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Beyond Fuzzer Islands: CPU Fuzzing via Smart Coordination
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessBeyond Fuzzer Islands: CPU Fuzzing via Smart Coordination