Wei Cui

About

Research Areas:

1. Efficient inference and cost-effective training on heterogeneous accelerators.

2. Acceleration of large language models and relational database systems.

3. Distributed and large-scale deep neural network training and query processing.

4. Accelerator virtualization and disaggregation.

Open Source Tools:

microsoft/tutel: Tutel MoE: An Optimized Mixture-of-Experts Implementation (github.com) (opens in new tab)

microsoft/antares: Antares: an automatic engine for multi-platform kernel generation and optimization (github.com) (opens in new tab)