What will you be responsible for?
Specifically, we are looking for an experienced algorithm engineer with a solid understanding of machine learning, deep learning, TensorFlow along with a background knowledge of image processing or signal processing algorithms. This role is to enable our R&D groups to evaluate various parallel HW architectures (such as GPUs) and explore & design algorithms to ensure that the most critical KLA AI algorithms can run optimally. This is a technical hands-on and lead role in research and implementation of these techniques and delivering production quality code. In a nutshell the position requires a strong and versatile background in algorithmic thinking, advanced deep learning tuning, and research in optimization methods. The position also requires a person with significant communication and technical leadership skills and the ability to navigate from relatively high-level requirements to low level algorithmic models. This position also requires to technically mentor new and junior team members to come up to speed and contribute effectively in on-going projects.
What we would like to see?
Working knowledge of AI algorithms, Machine Vision and ability to formulate problems into mathematical equations. Proficiency in TensorFlow, deep learning model training optimization, computational methods, C/C++ and Python is needed.
Minimum Education: BS/B. Tech in Computer Science/Computational Math/Physics. Masters and Ph.D. preferred. Need 3 to 8 years of work experience in related fields.
Finally, the candidate must be self-driven, curious and have the ability to navigate successfully with global teams including interacting with senior technical SW + HPC architects.
Exposure to computational aspects of TensorFlow. Understanding of parallel computing loads in the areas of image & signal processing.
Any prior exposure in programming and optimization in CUDA will be a plus. Additionally, any experience in optimizing large scale signal or computer vision algorithms domain would also be a major plus.
Background in Modern C++ concepts (C++ 11 through C++ 17), parallel STL library would also be a way to stand out from the crowd.