北京招聘

  

Responsibilities:

  

1. Bring neural network models to production, mainly focus on model pruning/quantization, and optimizing NN operators to reduce inference time, memory consumption, and power consumption.

2. Work with our AI IP vendor on driver integration, verification, and testing.

3. Design and develop the deep learning inference engine and SDK for our SoC.

      

Requirements:

  

1. MS in Computer Science, Electronic Engineering or other related fields.

2. 3+ years’ experience on high performance computing/heterogeneous computing/parallel computing(CUDA, OpenCL, GLSL, or NEON/AVX/SSE), Extensive experience with OpenCL and NEON is preferred.

3. Solid understanding of computer architectures , especially CPU, GPU, NPU.

4. Strong programming skills in C/C++, and familiar with Python.

5. Experience in algorithm optimization on GPU/DSP/FPGA is preferred.

6. Knowledge of machine learning/deep learning, especially deep neural networks is preferred.

7. Experience in popular deep learning frameworks, Tensorflow, Caffe or Pytorch etc, especially MACE, ncnn, TensorflowLite and PaddlePaddle mobile is preferred.

8. Experience in IoT / embedded system (i.e. smart home, smart display, smart IP camera for surveillance, etc.)  is preferred.

9. Good oral and written communication skills in English.

                  

瓴盛科技有限公司于2018年5月成立。由北京建广资产管理有限公司、大唐电信科技股份有限公司、高通（中国）控股有限公司、北京智路资产管理有限公司共同出资，注册资本为29.8亿人民币。公司致力于成为世界级的芯片设计企业，依托业内领先的技术平台，快速融入产业生态，赋能本地技术创新，与合作伙伴共同成长。公司立足中国、面向全球市场，主要提供智能手机芯片组、多元化智能物联网解决方案等的设计、封装、测试、客户支持、销售、技术咨询等业务。公司理念是通过提供应用于智能终端的产品、服务和解决方案，满足市场和广大客户的期望值，实现股东和利益相关者价值***化，为万物互联的社会做出贡献。
      目前公司拥有员工约500名，在上海、北京、成都、香港均设有分支机构，上海立可芯半导体科技有限公司为瓴盛科技有限公司全资子公司。