Job responsibilities
1.Design and implement compilers that generate efficient code for AI / computing applications, taking into account the performance and programmability of specific architectures;
2.Integrate architecture-specific performance estimate models into compiler optimization strategies to optimize code generation;
3.Analyze performance bottlenecks and develop solutions to improve performance for AI workloads;
4.Develop and maintain tools to automate the optimization process, using knowledge of specific architectures;
5.Stay current with new developments in AI technology and incorporate them into the compiler design as appropriate;
6.Promote open-source development and contribute to open-source projects related to compiler optimization for AI systems.
Requirements
1.BS, MS or PhD degrees in Computer Science, Electrical Engineering, or a related field;
2.Project experience in compiler development, with a focus on optimizing for AI or computing workloads;
3.Strong understanding of the performance and programmability of specific architectures, such as x86, ARM, CUDA GPU, RISC-V CPU;
4.Experience integrating architecture-specific performance models into compiler optimization strategies;
5.Familiarity with LLVM or other modern compiler infrastructures;
6.Strong programming skills in C++, with experience in Python and other scripting languages;
7.Familiarity with AI frameworks, such as TensorFlow, PyTorch, is preferred but not critical;
8.Strong problem-solving skills and the ability to analyze and optimize complex code;
9.Excellent communication skills and the ability to work effectively in a team environment;
10.Demonstrated commitment to open-source development and contributions to open-source projects.
