Responsibilities
- 1.Optimize and deploy AI algorithms on CPUs, discrete GPUs, integrated GPUs and dedicated edge inference chips.
- 2.Tailor algorithms to the characteristics of the hardware and the task, optimizing for performance, numerical precision and more.
- 3.Work with AI algorithm engineers on model conversion and custom operator development.
- 4.Work with AI algorithm engineers on model quantization and improving inference quality.
- 5.Contribute to the architecture, workflows and productivity tooling of the algorithm platform to improve the efficiency and quality of cross-platform algorithm engineering.
Requirements
- 1.Bachelor's degree or above in computer science or a related field; familiar with processor architectures (e.g., CPU, GPU, FPGA).
- 2.Proficient in C, C++ and Python, with solid programming fundamentals, good coding style and sound work habits.
- 3.Working knowledge of CMake and Makefile syntax and usage.
- 4.Experience with deployment frameworks such as TensorRT, MNN or Core ML is a plus.
- 5.Experience tuning performance on ARM, Intel x86 or other CPUs, familiarity with the relevant assembly instructions and multi-core parallel programming, and a basic understanding of parallel computing techniques such as CUDA or vector instructions are a plus.
- 6.Development experience with AI compilers such as TVM, MLIR or XLA is a plus.
- 7.Experience building a pipeline SDK from scratch is a plus.
How to Apply
- Email:
- bixue@attrsense.com
- Phone:
- Ms. Bi15316055810 (also on WeChat)
- Phone:
- Ms. Wang15201772885 (also on WeChat)
