Qidong Su 蘇起冬

My first name is pronounced as /tɕʰi tʊŋ/.

I am a Computer Science PhD student in University of Toronto, advised by Gennady Pekhimenko. I am also working as a System Software Engineer at NVIDIA. I got my bachelor degree from Shanghai Jiao Tong University (ACM Class).

My research focuses on accelerating programs on modern hardware. Currently I am optimizing inference speed of large-scale models. I'm happy to discuss topics including (but not limited to) machine learning systems, compiler designs, parallel programming, etc.

I like travelling by train. You can find related videos in my bilibili space (Chinese). I am also an amateur in linguistics (phonetics, Chinese dialects, Japanese).

• [MLSys 2025] Seesaw: High-throughput LLM Inference via Model Re-sharding
  (Outstanding Paper Honorable Mention🥇)
  Qidong Su, Wei Zhao, Xin Li, Muralidhar Andoorveedu, Chenhao Jiang, Zhanda Zhu, Kevin Song, Christina Giannoula, Gennady Pekhimenko
  preprint slides
• [PACT 2024] BOOM: Use your Desktop to Accurately Predict the Performance of Large Deep Neural Networks
  Qidong Su, Jiacheng Yang, Gennady Pekhimenko
  preprint paper

• Email: qdsu ät cs.toronto.edu