Yudong Pan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Yudong Pan has a total of 30 co-authors. Here are the top co-authors by number of shared publications:
- Mengdi Wang
Chinese Academy of Sciences
5 shared publications - Ying Wang
Chinese Academy of Sciences
5 shared publications - Yinhe Han
Chinese Academy of Sciences
5 shared publications - Cangyuan Li
SKLP, Institute of Computing Technology, Chinese Academy of Sciences;, Beijing, Beijing, China
3 shared publications - Yiqi Liu
Institute of Computing Technology, CAS, Beijing, China and University of Chinese Academy of Sciences, Beijing, China
3 shared publications - Huawei Li
Chinese Academy of Sciences
2 shared publications
Latest Publications
- Open Access
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
TriMoE: Augmenting GPU with AMX-Enabled CPU and DIMM-NDP for High-Throughput MoE Inference via Offloading
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessTriMoE: Augmenting GPU with AMX-Enabled CPU and DIMM-NDP for High-Throughput MoE Inference via Offloading
- Open Access
- 22 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 1349-1365
Ouroboros: Wafer-Scale SRAM CIM with Token-Grained Pipelining for Large Language Model Inference
- 22 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 1349-1365
Open AccessOuroboros: Wafer-Scale SRAM CIM with Token-Grained Pipelining for Large Language Model Inference
- Open Access
- 21 October 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Design Automation of Electronic Systems
- Vol. 30 (6) , 1-21
AutoSilicon: Scaling Up RTL Design Generation Capability of Large Language Models
Hardware description language (HDL) code designing is a critical component of the chip design process, requiring substantial engineering and time resources. Recent advancements in large language models (LLMs), such as GPT series, have shown promise in automating HDL code generation. However, current LLM-based approaches face significant challenges in meeting real-world hardware design requirements, particularly in handling complex designs and ensuring code correctness. Our evaluations reveal that the functional correctness rate of LLM-generated HDL code significantly decreases as design complexity increases. In this article, we propose the AutoSilicon framework, which aims to scale up the hardware design capability of LLMs. AutoSilicon incorporates an agent system, which (1) allows for the decomposition of large-scale, complex code design tasks into smaller, simpler tasks; (2) provides a compilation and simulation environment that enables LLMs to compile and test each piece of code it generates; and (3) introduces a series of optimization strategies. Experimental results demonstrate that AutoSilicon can scale hardware designs to projects with code equivalent to over 10,000 tokens. In terms of design quality, it further improves the syntax correctness rate and functional correctness rate compared with approaches that do not employ any extensions. For example, compared to directly generating HDL code using GPT-4-turbo, AutoSilicon enhances the syntax correctness rate by an average of 35.8% and improves functional correctness by an average of 35.6%.
- 21 October 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Design Automation of Electronic Systems
- Vol. 30 (6) , 1-21
Open AccessAutoSilicon: Scaling Up RTL Design Generation Capability of Large Language Models
Hardware description language (HDL) code designing is a critical component of the chip design process, requiring substantial engineering and time resources. Recent advancements in large language models (LLMs), such as GPT series, have shown promise in automating HDL code generation. However, current LLM-based approaches face significant challenges in meeting real-world hardware design requirements, particularly in handling complex designs and ensuring code correctness. Our evaluations reveal that the functional correctness rate of LLM-generated HDL code significantly decreases as design complexity increases. In this article, we propose the AutoSilicon framework, which aims to scale up the hardware design capability of LLMs. AutoSilicon incorporates an agent system, which (1) allows for the decomposition of large-scale, complex code design tasks into smaller, simpler tasks; (2) provides a compilation and simulation environment that enables LLMs to compile and test each piece of code it generates; and (3) introduces a series of optimization strategies. Experimental results demonstrate that AutoSilicon can scale hardware designs to projects with code equivalent to over 10,000 tokens. In terms of design quality, it further improves the syntax correctness rate and functional correctness rate compared with approaches that do not employ any extensions. For example, compared to directly generating HDL code using GPT-4-turbo, AutoSilicon enhances the syntax correctness rate by an average of 35.8% and improves functional correctness by an average of 35.6%.
- Open Access
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 131-146
COMET: Towards Practical W4A4KV4 LLMs Serving
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 131-146
Open AccessCOMET: Towards Practical W4A4KV4 LLMs Serving
- Top Cited PapersOpen Access
- 23 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Data is all you need: Finetuning LLMs for Chip Design via an Automated design-data augmentation framework
Recent advances in large language models have demonstrated their potential for automated generation of hardware description language (HDL) code from high-level prompts. Researchers have utilized fine-tuning to enhance the ability of these large language models (LLMs) in the field of Chip Design. However, the lack of Verilog data hinders further improvement in the quality of Verilog generation by LLMs. Additionally, the absence of a Verilog and electronic design automation (EDA) script data augmentation framework significantly increases the time required to prepare the training dataset for LLM trainers. This paper proposes an automated design-data augmentation framework, which generates high-volume and high-quality natural language aligned with Verilog and EDA scripts. For Verilog generation, it translates Verilog files to an abstract syntax tree and then maps nodes to natural language with a predefined template. For Verilog repair, it uses predefined rules to generate the wrong verilog file and then pairs EDA Tool feedback with the right and wrong verilog file. For EDA Script generation, it uses existing LLM(GPT-3.5) to obtain the description of the Script. To evaluate the effectiveness of our data augmentation method, we finetune Llama2--13B and Llama2-7B models using the dataset generated by our augmentation framework. The results demonstrate a significant improvement in the Verilog generation tasks with LLMs. Moreover, the accuracy of Verilog generation surpasses that of the current state-of-the-art open-source Verilog generation model, increasing from 58.8% to 70.6% with the same benchmark. Our 13B model (ChipGPT-FT1) has a pass rate improvement compared with GPT-3.5 in Verilog generation and outperforms in EDA script (i.e., SiliconCompiler) generation with only 200 EDA script data.
- 23 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Top Cited PapersOpen AccessData is all you need: Finetuning LLMs for Chip Design via an Automated design-data augmentation framework
Recent advances in large language models have demonstrated their potential for automated generation of hardware description language (HDL) code from high-level prompts. Researchers have utilized fine-tuning to enhance the ability of these large language models (LLMs) in the field of Chip Design. However, the lack of Verilog data hinders further improvement in the quality of Verilog generation by LLMs. Additionally, the absence of a Verilog and electronic design automation (EDA) script data augmentation framework significantly increases the time required to prepare the training dataset for LLM trainers. This paper proposes an automated design-data augmentation framework, which generates high-volume and high-quality natural language aligned with Verilog and EDA scripts. For Verilog generation, it translates Verilog files to an abstract syntax tree and then maps nodes to natural language with a predefined template. For Verilog repair, it uses predefined rules to generate the wrong verilog file and then pairs EDA Tool feedback with the right and wrong verilog file. For EDA Script generation, it uses existing LLM(GPT-3.5) to obtain the description of the Script. To evaluate the effectiveness of our data augmentation method, we finetune Llama2--13B and Llama2-7B models using the dataset generated by our augmentation framework. The results demonstrate a significant improvement in the Verilog generation tasks with LLMs. Moreover, the accuracy of Verilog generation surpasses that of the current state-of-the-art open-source Verilog generation model, increasing from 58.8% to 70.6% with the same benchmark. Our 13B model (ChipGPT-FT1) has a pass rate improvement compared with GPT-3.5 in Verilog generation and outperforms in EDA script (i.e., SiliconCompiler) generation with only 200 EDA script data.