mini-DeepInfer
An inference runtime, from request to next token.
A readable inference pipeline for exploring request scheduling, model execution, and cache management.
Hi, I'm Chenxi Ma. I'm currently interning at UW Systems Lab under the guidance of Baris Kasikci.
My interests lie in LLM inference optimization and GPU systems. I explore how model capabilities connect with the systems that execute them.
My technical interests include inference runtimes, GPU scheduling, mixture-of-experts models, and communication and computation in distributed inference. I'm also interested in co-design across model architectures, agent harnesses, hardware, and infrastructure.
This page brings together my projects and research interests. Technical articles and experiment notes live on my separate blog.
Research Interests
Exploring efficient model execution and reliable reasoning and decision-making.
- Efficient Inference
- LLM inference, runtime design, KV caches, and request scheduling.
- GPU & Distributed Systems
- GPU programming, MoE execution, and communication–computation co-design.