Youwei Xiao
Youwei Xiao 肖有为
School of Integrated Circuits
Peking University
Beijing, China
Final-year Ph.D., expected 2027
I am a final-year Ph.D. candidate at the School of Integrated Circuits, Peking University, advised by Prof. Yun Liang, and a Large Language Model Algorithm Intern at ByteDance Seed. My research pursues compiler-driven hardware-software co-design for agile chips.
I am expected to graduate in 2027 and am open to industry positions (especially top-talent programs). If you are aware of suitable or well-matched positions, please feel free to contact me via email.
My current work creates an extensible benchmark of kernel programming and optimization tasks for dataflow-architecture chips, and builds the layered knowledge, data collection, and training workflows that let models keep improving. The goal is a methodology and workflow that enable models and agents to rapidly acquire expert-level kernel optimization capabilities for new chip architectures.
Architecture description and exploration. I combine compiler analysis, formal methods, and architecture DSLs to represent and search design spaces, and to connect customized hardware to programmable software.
- ISAMORE (ASPLOS 2026 Best Paper) discovers reusable custom instructions from equivalent program fragments.
- ScaleISA (ASPLOS 2027) extends that line to MLIR, identifying scalable instructions by multi-level e-graph anti-unification.
- Cayman identifies accelerator opportunities in complete applications while co-optimizing control flow and data access.
- APS/Aquas describes customized architectures through DSLs and exposes them through complete hardware/software compiler stacks.
Chip DSL and compilation. I design DSLs and intermediate representations at multiple abstraction levels, then build compilation and synthesis passes for timing, microarchitecture, implementation selection, mapping, and scheduling.
- Hector provides a multi-level MLIR foundation for hardware synthesis.
- Cement couples the cycle-deterministic CmtHDL DSL with the CmtC compiler for timing analysis and control synthesis.
- Clay and SkyEgg automate microarchitecture-aware implementation selection and scheduling.
Compiler optimization and LLM systems.
- EggMind synthesizes equality-saturation strategies with LLM guidance.
- IntelliC studies inspectable compiler representations for human-model collaboration.
- Spine organizes verification-bounded co-synthesis across design intent, architecture, compiler, hardware, and execution evidence.
- PTO Runtime supports dynamic kernel fusion and distributed execution of compiled task graphs on Ascend chips and LingQu SuperPods.
- Hive provides infrastructure for multi-agent inference workloads.
Academic service. I serve on the Student Technical Program Committee of MICRO 2026 and the Technical Program Committee of MLCAD 2026. I also reviewed journal submissions for IEEE TCAD and ACM TECS in 2024.
news
| Sep 23, 2026 | Our paper ScaleISA has been accepted to ASPLOS 2027. It is the MLIR extension of ISAMORE, using multi-level e-graph anti-unification to customize scalable instructions. |
|---|---|
| Aug 12, 2026 | Our paper Aquas has been accepted to ICCAD 2026. It presents holistic hardware-software co-optimization based on MLIR. |
| May 28, 2026 | Awarded 博士研究生校长奖学金 for the 2026-2027 academic year, with 10 recipients selected from the School of Integrated Circuits. |
selected publications
- ICCADAquas: Enhancing Domain Specialization through Holistic Hardware-Software Co-Optimization based on MLIRIn Proceedings of the 45th IEEE/ACM International Conference on Computer-Aided Design (ICCAD ’26), 2026