1 month ago
Taipei, TaiwanSenior
Responsibilities
- Optimize TensorFlow, PyTorch, SGLang, and other deep learning frameworks on AMD GPUs through upstream open-source contributions.
- Profile and tune large-scale training and inference models for performance on AMD hardware.
- Design, implement, and optimize GPU kernels using HIP, Triton, or other relevant tools.
- Collaborate with GPU library and compiler teams to align kernel optimizations with full-stack performance goals.
- Support optimization, feature development, and scaling of the SGLang LLM framework across AMD GPU platforms.
- Tune performance across multi-GPU and multi-node environments, including inference parallelism and collective communication.
- Integrate and optimize runtime execution through graph compilers such as XLA, TorchDynamo, or custom pipelines.
- Partner with external maintainers and contribute performance improvements to open-source projects.
- Apply debugging, profiling, test-driven development, and continuous integration practices.
Requirements
- Bachelor’s and/or master’s degree in computer science, computer engineering, electrical engineering, or a related field.
- Strong C++ and/or Python programming experience in Linux environments.
- Demonstrated ability to debug, profile, and optimize performance-critical code.
- Hands-on experience with SGLang or similar LLM inference frameworks is highly preferred.
- Background in compiler design or familiarity with LLVM, MLIR, or ROCm is preferred.
- Experience running and scaling workloads on large heterogeneous CPU and GPU clusters using distributed training or inference strategies.
- Experience contributing to or integrating optimizations into deep learning frameworks such as PyTorch or TensorFlow.
- Strong knowledge of HIP, CUDA, or other GPU programming models; GCN or CDNA architecture experience is preferred.
Benefits
- AMD benefits are described in the AMD benefits at a glance materials.
Categories
About AMD
AMD designs and sells CPUs, GPUs, and adaptive/embedded computing products for PCs, data centers, gaming, and edge devices. Its portfolio includes Ryzen and EPYC processors, Radeon and Instinct graphics, and adaptive SoCs from its Xilinx acquisition, sold to OEMs, cloud providers, and device makers. Founded in 1969 and headquartered in Santa Clara, it is a public company on NASDAQ and supplies semi-custom chips for major game consoles.
