Ethan Che
Software Engineer · PyTorch · Meta Superintelligence Labs
I’m a software engineer on the PyTorch team within Meta Superintelligence Labs.
I work on ML compilers and kernel development for heterogeneous hardware platforms (GPU, TPU).
Selected work
Helion on TPU: Towards Hardware Heterogeneous Kernel Authoring
Dunfan Lu, Yifei Xu, Jongsok Choi, Ethan Che, Oguz Ulgen, Jason Ansel, Yarong Mu, Theotime Combes, Emilio Cota, Freya Azad
Towards Free Normalization: Fusing Normalization into GEMM and Attention Kernels
Jacky (Junqing) Zhou, Hongtao Yu, Jackie (Jiaqi) Xu, Menglu Yu, Ethan Che, Han Xu, Darren Liu, Peng Chen (Dev Infra), Daohang Shi, Max Leung
Accelerating Autotuning in Helion with Bayesian Optimization
Ethan Che, Oguz Ulgen, Max Balandat, Jongsok Choi, Jason Ansel
Exploration Sizing via Model-Predictive Control
Ethan Che, Hakan Ceylan, James McInerney, and Nathan Kallus
How Well do LLMs Compress Their Own Chain-of-Thought? A Token Complexity Approach
Ayeong Lee*, Ethan Che*, and Tianyi Peng
Stochastic Gradient Descent with Adaptive Data
Ethan Che, Jing Dong, and Xin T. Tong
Differentiable Discrete Event Simulation for Queuing Network Control
Ethan Che, Jing Dong, and Hongseok Namkoong
* Equal contribution.