工作內容
1. Design and develop an NPU compiler that translates high-level neural network graphs (e.g. ONNX, TensorFlow, PyTorch) into optimized machine code for NPU execution binaries
2. Perform computational graph optimizations, including operator fusion, kernel selection, and memory layout transformation to optimize inference latency, power efficiency, and memory footprint
3. Profile compiled workloads to identify performance bottlenecks and optimization opportunities
4. Integrate compiler with runtime libraries, drivers, and software stack
工作說明
-
工作縣市:臺北市
- 上班地點:台北市內湖區
-
工作待遇:月薪80,000~250,000元
-
上班時段:依公司規定
-
需求人數:不限
條件要求
-
工作經歷:
經歷不拘
-
學歷要求:碩士
-
科系要求:
無填寫
-
專長需求:
-
擅長工具:
- 具備駕照:
-
其他條件:
https://www.104.com.tw/job/91u2w