GPU Kernel Developer - AI Models

As a GPU Kernel Developer specializing in AI models at AMD, your primary responsibility is to develop high-performance GPU kernels for cutting-edge and upcoming GPU hardware. You will collaborate with a team of industry experts, leveraging the latest hardware and software technologies to drive innovation in the field. To excel in this role, you must possess significant experience in GPU kernel development and optimization for AI/HPC applications. Your expertise should include a deep understanding of GPU computing, hardware architecture, and proficiency in HIP, CUDA, OpenCL, and Triton development. Effective communication skills are essential as you will be required to work within a team environment and convey complex technical concepts to both technical and non-technical audiences. Key responsibilities include developing high-performance GPU kernels for essential AI operators on AMD GPUs, optimizing GPU code through structured methodologies, and supporting critical workloads in NLP/LLM, Recommendation, Vision, and Audio domains. You will collaborate closely with system-level performance architects, GPU hardware specialists, and various validation and marketing teams to analyze and enhance training and inference processes for AI applications. Furthermore, you will engage with open-source framework maintainers to align with their requirements and integrate code changes upstream, debug, maintain, and optimize GPU kernels, and drive AI operator performance improvements. Your expertise in software engineering best practices will be crucial in ensuring the quality and efficiency of the developed solutions. Preferred qualifications for this role include knowledge of GPU computing technologies such as HIP, CUDA, OpenCL, and Triton, experience in optimizing GPU kernels, proficiency in profiling and debugging tools, a strong foundation in GPU hardware, and excellent programming skills in C/C++/Python. Additionally, a Master's or PhD in Computer Science, Computer Engineering, or a related field is preferred. Join us at AMD, where we are dedicated to pushing the boundaries of innovation and solving the world's most significant challenges through transformative technology.,