Skip to content
View Phil-amd's full-sized avatar
🎯
Focusing
🎯
Focusing
0 followers · 7 following
  • AMD

Achievements

Achievement: Pair Extraordinaire

Achievements

Achievement: Pair Extraordinaire

Organizations

@ROCm

Block or report Phil-amd

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Phil-amd/README.md

Phil Li

Compiler engineer at AMD, working on GPU compiler infrastructure and the kernel stack underlying modern AI workloads.

  tensor program
        │  lower
        ▼
   MLIR dialects ──► LLVM IR ──► AMDGPU ISA ──► the machine
        ▲                                            │
        └──────────── read the disassembly ◄─────────┘

My work sits at the boundary where a high-level tensor program is lowered into machine code that fully utilizes the hardware: IR design, code generation quality, and the domain-specific languages that allow kernel authors to express layout and scheduling without resorting to raw intrinsics. When a kernel misses peak, the answer is in the generated assembly — so that is where I start.

Areas of Interest

  • LLVM and MLIR compiler infrastructure
  • Code generation and performance analysis for AMD CDNA GPUs
  • Kernel DSLs and layout abstractions for AI workloads
  • Large language model inference and serving performance on ROCm

Current Work

  • AMD LLVM — AMDGPU backend and the ROCm compiler toolchain
  • FlyDSL — Flexible LaYout DSL, a Python front end for authoring high-performance GPU kernels
  • aiter — AI Tensor Engine for ROCm

Technologies

LLVM · MLIR · ROCm / HIP · C++ · Python · Triton

Every abstraction leaks. The interesting work is deciding where.

Pinned Loading

  1. sgl-project/sglang sgl-project/sglang Public

    SGLang is a high-performance serving framework for large language models and multimodal models.

    Python 31.4k 7.7k

  2. ROCm/aiter ROCm/aiter Public

    AI Tensor Engine for ROCm

    Python 522 455

  3. ROCm/FlyDSL ROCm/FlyDSL Public

    FlyDSL is the Python front‑end of the project: a Flexible Layout Python DSL for expressing tiling, partitioning, data movement, and kernel structure at a high level.

    Python 257 103

  4. llvm-project llvm-project Public

    Forked from ROCm/llvm-project

    This is the AMD-maintained fork of the LLVM git repository. This repository accepts pull requests and issues related to AMD fork-specific topics (amd/*). For all other issues/PRs, please submit ups…

    LLVM