文章主题:【Intel 北京实习】Machine Learning Workload Development Engineer
发信人: zfhr, 信区: 招聘求职
标 题: Re: 【Intel 北京实习】Machine Learning Workload Development Engineer
发信站: 斗转星云 BBS站 (2020/6/24 10:30:38)
【Intel 北京实习】Machine Learning Workload Development Engineer
【我们会为你提供这样的工作】 Intel's Visual Technology Team (VTT) develops best in class compute technology that is a critical part of our major product lines. We are looking for Intern engineers to develop tensor graph compiler targeting machine learning workloads for better utilizing the Gen compute architecture. The machine learning workload team in VTT is responsible for machine learning workload development and performance analysis. We support pre-silicon analysis and projection, Gen architectural technical readiness (TR) support, what-if analysis in TR and beyond.
Responsibilities: Responsibilities will include, but are not limited to: ? Contribute to the kernel generation to execute tensor graph on Gen architecture ? Develop parameterized machine learning kernels running on Gen architecture ? Maintain and develop code in the regression environment ? Analyze, validate, and report out kernel performance of new Gen compute architecture in pre-Si model
Minimum Qualifications ? Enrolled and working toward PhD Degree in in Computer Science, Computer Engineering, Electrical Engineering, or a related field. ? Programming skills: specifically, python, C++; GPU parallel programming. ? Data analysis, debugging and problem solving skills/experience. ? Experience explaining complex technical concepts. Preferred Qualifications An ideal candidate will have knowledge and/or experience in some of the following: ? Deep learning workload understanding, e.g. CNN, RNN etc. ? Machine learning kernel development experience in GPU. ? Understanding compiler and programming language design is a plus ? Tensor compiler experience is a big plus. ? Excellent understanding of GPU compute architecture and technology