Computational Research Scientist - High Performance Computing
Organisation Name: Lawrence Berkeley National Laboratory
Organisation Type:
City:
State:
Country:
Job Description:
Berkeley Lab’s Computational Research Division has an opening for a Research Scientist to conduct fundamental and applied research in the areas of high-performance computing, parallel computing, and performance modeling. The successful candidate will join the Performance and Algorithms (PAR) group that is funded to work on high-impact projects on Exascale computing, Roofline and performance analysis, computational biology, and computations on graphs and sparse matrices. What You Will Do: Develop and extend the Department of Energy's performance modeling capabilities to capture the key performance facets of scientific computing, graph analytic, and deep learning applications running on current and future multicore- and accelerator-based supercomputers. Collaborate with domain scientists and applied mathematicians to identify computational bottlenecks in multicore- and accelerator-based distributed memory applications, implement solutions, and evaluate the resultant performance. Publish and present research results in high-impact journals and top conferences in the areas of high-performance computing, parallel computing, and scientific computing. Support supervision of postdocs and students working in the areas of high-performance computing, parallel computing, and performance modeling. Write research proposals and reports. Engagement with broader scientific community, societies, committees, and councils. What is Required: PhD in Computer Science or a related field. Experience in designing and implementing GPU-accelerated software for the fields of scientific computing, graph algorithms, or deep learning. Proficiency in C and/or C++ programming language. Proficiency in one or more of the parallel libraries/languages such as MPI, OpenMP, CUDA. Demonstrated ability to design and implement performance modeling and performance analysis capabilities on multicore and accelerated distributed-memory computing platforms. Demonstrated ability to analyze multicore and accelerated distributed-memory applications, design and implement optimizations, and convey actionable feedback to computational scientists, applied mathematicians, and computer architects. Demonstrated ability to publish in top journals and conferences. Desired Qualifications: Software engineering tools such as make and git. One or more of the popular scripting languages such as Python, Julia, or R. The posting shall remain open until the position is filled. Notes: This is a full-time 2-year, career-track term appointment that may be renewed to a maximum of five years and that may be converted to career based upon satisfactory job performance, continuing availability of funds, and ongoing operational needs. This position may be subject to a background check. Any convictions will be evaluated to determine if they directly relate to the responsibilities and requirements of the position. Having a conviction history will not automatically disqualify an applicant from being considered for employment. Work will be primarily performed at Lawrence Berkeley National Lab, 1 Cyclotron Road, Berkeley, CA.
