1 day ago
Taipei, TaiwanSenior
Responsibilities
- Optimize TensorFlow, PyTorch, and SGLang performance on AMD GPUs through framework and open-source contributions.
- Develop, profile, tune, and optimize large-scale deep learning training and inference models.
- Design and optimize GPU kernels using HIP, Triton, or related GPU programming tools.
- Collaborate with GPU library and compiler teams to integrate kernel-level optimizations.
- Improve SGLang features, scaling, and LLM inference performance across AMD GPU platforms.
- Tune distributed performance across multi-GPU and multi-node systems, including inference parallelism and collective communication.
- Integrate graph compiler and runtime execution technologies such as XLA and TorchDynamo.
- Work with open-source maintainers and contribute optimizations upstream.
- Apply debugging, profiling, test-driven development, and CI practices.
Requirements
- Strong experience developing in C++ and/or Python 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.
- Knowledge of compiler design or technologies such as LLVM and MLIR is a plus.
- Experience scaling distributed training or inference workloads on heterogeneous CPU/GPU clusters.
- Experience integrating or contributing optimizations to PyTorch, TensorFlow, or similar deep learning frameworks.
- Strong knowledge of HIP, CUDA, or other GPU programming models; GCN or CDNA experience is preferred.
- Bachelor’s and/or master’s degree in computer science, computer engineering, electrical engineering, or a related field.
Benefits
- Hybrid work arrangement.
- AMD benefits are offered, as described in the company’s benefits overview.
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.