$80K — $100K *
The National Center for Computational Sciences (NCCS) provides state-of-the-art computational and data science infrastructure for technical and scientific professionals to accelerate scientific discovery and engineering advances across a broad range of disciplines. As an important part of the broader High-Performance Computing (HPC) infrastructure, the division also hosts the Oak Ridge Leadership Computing Facility (OLCF), a Department of Energy Office of Science User Facility.
We seek to hire a Computer Scientist, Applied Mathematician, or Computational Scientist, specializing in performance analysis and optimization of HPC software to join our Algorithms and Performance Analysis Group within the Science Engagement section of NCCS. In this role, you will serve as a liaison between NCCS and the users of the NCCS computing resources and relevant NCCS vendor partners, providing deep technical expertise in high performance computing that can be applied to a diverse range of scientific disciplines.
Members of the Algorithms and Performance Analysis Group within the Scientific Engagement Section partner with NCCS users across all computational domains to enable, promote and optimize the use of, and to maximize the benefits of using the world class facilities provided by NCCS. Members of the group liaise with and represent users and user applications within NCCS, carry out independent research, and provide broad expertise, technical input and service within the division to assist in strategic decision making, future system and infrastructure development, procurement and deployment, and to further the mission of NCCS and Oak Ridge National Laboratory.
• Represent the needs of a broad range of application communities for NCCS planning and decision-making.
• Forge collaborations with other researchers, both internally at ORNL and externally, on topics of mutual interest. This includes collaborating scientifically with awarded OLCF projects.
• Participate in the development of proposals for computational or computer science research projects.
• Conduct research and report results in journal articles, conference papers, and technical manuals.
• Collaborate with other computational and computer scientists to develop, integrate, port, tune, and debug applications to take advantage of the capabilities of the OLCF leadership-computing environment and other NCCS resources.
• Assist users in moving data and programs into and out of the NCCS environment and help with scheduling applications at scale to meet the scientific needs of the users.
• Provide expertise and leadership in current OLCF procurement activities and assist in capability planning exercises for future OLCF procurements.
• Advise and work with the NCCS Systems and Operations sections to understand the needs of computational scientists in a broad range of domains. Participate directly in formulating policies to maximize scientific productivity of the NCCS compute resources.
• Participate in community-wide efforts to define the requirements for system software, libraries, and tools to meet the needs of the application users.
• A Ph.D. in computer science, applied mathematics, or other related field (e.g. astrophysics, biophysics, chemistry, climate science, engineering, materials science, nuclear and particle physics, plasma physics, etc.).
• Minimum of 2 years of experience in computer science, applied mathematics, or other related field with demonstrated experience developing and/or tuning application codes and/or libraries for high performance computing resources.
• Experience with numerical methods, parallel algorithms, MPI, and a common scientific computing programming language (i.e., Fortran, C++ and/or C).
• Parallel software development targeting large-scale computational resources.
• The ability to work in a dynamic, team environment.
• Excellent interpersonal skills, oral and written communication skills, organizational skills, and strong personal motivation.
• 6+ years of relevant research experience beyond a Ph.D.
• Experience with one or more prominent simulation codes or libraries that use the Leadership Computing Facilities.
• Experience in developing software for heterogeneous, accelerated computer architectures (for example using GPUs) using approaches like CUDA/HIP, compiler directives like OpenMP or OpenACC, or abstractions such as Kokkos.
• Experience using performance tools such as TAU, HPCToolkit, MAP, CrayPAT to analyze and optimize multi-node parallel HPC software.
• Experience using performance tools such as NVIDIA's Nsight and AMD's rocProf to analyze and optimize software running on accelerators (e.g., GPUs).
• Experience in a domain aligned with US DOE Office of Science mission areas.
• Experience with multiple common scientific computing programming languages and programming models.
• Experience with computing at scale (for example, computing on a resource in the top 20 of the Top500 list) using MPI and/or one-sided asynchronous models.
• Knowledge of the software stack on a leadership computing platform, including software development toolchain, runtime systems, and debugging/performance tools.
Benefits at ORNL
UT Battelle offers an exceptional benefits package to include matching 401K, Pension Plan, Paid Vacation and Medical/Dental plan. Onsite amenities include Credit Union, Medical Clinic, and free Fitness facilities.
Valid through: 4/23/2021
$200K — $250K
13 days ago