Performance Modeling for Large-Scale AI Workloads

Research questions

• How much execution detail does a model need to answer a particular design question?
• How can we validate a lightweight model for large-scale AI workloads?
• How do parallelism and interconnect choices affect multi-GPU workload performance?
• When are sampled simulation or regression-based predictions sufficient for an architecture study?

Publications

2026

arXiv

ArchSim: Computer Architecture Simulation as a Service

Sabila Al Jannat, Wenhan Lyu, Le Khanh Trinh Mai, Huizhi Zhao, Zhuoyan Zheng, Katherine E. Isaacs, Yifan Sun

PDF

arXiv

Akita: A High Usability Simulation Framework for Computer Architecture

Sabila Al Jannat, Ying Li, Mengyang He, Xuzhong Wang, Huizhi Zhao, Jingxiang Sun, Daoxuan Xu, Enze Xu, Yifan Sun

PDF

GPGPU 2026

Did You Win the GPU Cloud Lottery? Benchmarking from TFLOPS to Tokens/$

Platon Slynko, Jay Lu, Jason Cornick, Laksh Sharma, Shou-Kai Cheng, Benjamin Cornick, Arthur O. Dias dos Santos, Caique Sobral, Carmen Li, Daoxuan Xu, Xinxin Mei, Yifan Sun

PDF

2025

ISCA 2025

TrioSim: A Lightweight Simulator for Large-Scale DNN Workloads on Multi-GPU Systems

Ying Li, Yuhui Bao, Gongyu Wang, Xinxin Mei, Pranav Vaid, Anandaroop Ghosh, Adwait Jog, Darius Bunandar, Ajay Joshi, Yifan Sun

PDF

Slides

2023

MICRO 2023

Path Forward Beyond Simulators: Fast and Accurate GPU Execution Time Prediction for DNN Workloads

Ying Li, Yifan Sun, Adwait Jog

PDF

MICRO 2023

Photon: A Fine-grained Sampled Simulation Methodology for GPU Workloads

Changxi Liu, Yifan Sun, Trevor E. Carlson

PDF

ISPASS 2023

A Regression-based Model for End-to-End Latency Prediction for DNN Execution on GPUs

Ying Li, Yifan Sun, Adwait Jog

PDF

2022

PACT 2022

NaviSim: A Highly Accurate GPU Simulator for AMD RDNA GPUs

Yuhui Bao, Yifan Sun, Zlatan Feric, Michael Tian Shen, Micah Weston, José L. Abellán, Trinayan Baruah, John Kim, Ajay Joshi, David Kaeli

PDF

Related software

• Akita
• MGPUSim
• TrioSim
• Lab software