平头哥平头哥-AI 框架技术专家-杭州

任职要求

1. 硕士以上，计算机科学相关专业， 至少二年/五年工作经验；
2. 熟悉C/C++编程，熟悉Python编程，有Unix/Linux环境下开发经验；
3. 熟悉一种或几种主流机器学习框架（如TensorFlow/Pytorch等）的工作原理；
4. 熟悉从框架层进行模型性能与精度的量化分析
拥有以下任何方面的经验者优先：
a. 熟悉机器学习相关领域的模型算法结构与原理包括：计算机视觉, 视频处理与解析, 图片处理, 语音识别, 自然语言理解；
b. 有并行编程经验，如CUDA/OpenCL/OpenMP等；
c. 有编译器开发经验；
5. 良好的英语听说读写能力；
6. 沟通表达能力良好，富有事业心和团队合作精神。

1. Master or higher degree in CS/EE/Math or equivalent is required with minimum 2/5 years working experience.
2. Strong C/C++ and Python programming skills, and Unix/Linux development experience is required as well.
3. Experience on Machine-Learning and Neural Network is required, and at least being familiar with one major machine-learning framework(like Tensorflow, Pytorch, etc.).
4. Be familiar with model performance profiling and accuracy analysis from the framework layer
Having these experience is a good plus:
a. Experience with any of following areas: computer vision, video processing and analysis, image processing, speech recognition, NLP.
b. Experience on parallel programming such as CUDA/OpenCL/OpenMP etc.
c. Experience on compiler development.
5. Good oral, reading and writing skills in English.
6. Excellent communication skill, innovation and teamwork are also required.

工作职责

1. 在AI 芯片上移植各类开源机器学习框架，分析解决移植过程中的编译与运行时问题；
2. 分析和验证各类机器学习算法在AI芯片上的功能、性能与精度，从框架层分析并解决发现的功能、精度与性能问题，为芯片设计与AI软件栈提供参考；
3. 对模型、框架、AI生态兼容做竞品对比、原理分析，为AI芯片客户提供问题支持和解决方案。

1. Enable open source machine learning frameworks on AI chips to analyze and solve all kinds of problems in the porting process.
2. Analyze and verify the functions, performance and accuracy of various machine learning algorithms on AI chips, Analyze and solve the problems from the framework layer, and delivering reference data for chip design and AI software stack.
3. Compare and analyze the competitive products from the view of open source frameworks and AI ecological compatibility to provide customers  support and solutions.