Be part of something great!
The High Performance Computing Services group is looking for a HPC Scientific User Services Engineer to provide high performance computational resources and related support services to the Lawrence Berkeley National Laboratory and University of California researchers. This position is responsible for problem management and consulting, helping with user code optimization, debugging and scaling, strategic user project support, documentation, training, and third-party applications and library support. Additional responsibilities include implementing and managing job resource management, scheduler, and user allocation software as well as assisting with the benchmarking, integration, and acceptance testing of new Linux High Performance Computing (HPC) clusters.
Specifically you will:
- Provide broad spectrum of HPC user services support to LBNL and UC researchers. This includes user assistance porting, compiling, optimizing, debugging, and running parallel applications on Linux clusters; and performance tuning, profiling, and benchmarking of applications.
- Analyze the mix of user job requirements and implement job queues and scheduler algorithms to load balance and make efficient use of computational resources while satisfying user requirements for turnaround time.
- Develop and maintain scripts to manage user account allocations, track usage, and provide reports.
- Create and maintain user level documentation on use of systems.
- Keep users informed of software and system changes.
- Provide training to help users transition to new technologies and use resources effectively.
- Write online tutorials.
- Prepare and present classroom and teleconference lectures.
- Keep abreast of developments in the high performance computing field.
- Build and implement third party software to support operating system, high performance interconnects, variants of MPI, and scientific applications and libraries.
- Keep such software up-to-date, understand the users' software needs, and help evaluate requests for new software.
- Test systems and software before and after major system upgrades. Run benchmarks as needed.
- Scientific background having parallel programming, debugging and scaling experience.
- Background in data analysis.
- The candidate selected will have major responsibilities for helping users select the right software and algorithms for their codes, debug their codes with a parallel debugger, tune codes for parallel performance, and analyze large datasets.
QUALIFICATIONS:
- This position requires the candidate to have an advanced degree in Computer Science or scientific field; a minimum of 8 years scientific computing experience, including a minimum of 2 years experience in providing user support for high performance computing Linux clusters.
- Experience in implementing and maintaining various job schedulers and scheduling algorithms to load balance user jobs on systems.
- Software development experience and related skills are essential in order to provide expert user assistance with their applications or user environment.
- Must be able to demonstrate parallel programming proficiency in a procedural language such as C or Fortran and scripting languages such as PERL or Python.
- Experience with popular compilers (e.g. GCC, Portland Group, Intel, Pathscale), parallel programming models (MPI), effective use of multi-core nodes, debugging and performance tuning techniques and tools, and scientific libraries.
- Experience doing benchmarking activities to test or accept systems.
- Ability to build, package, install, and test 3rd party commercial or public domain scientific software.
- Ability to plan, organize, prioritize many tasks, and complete assigned tasks and projects with minimal supervision and must have demonstrated ability to assume lead role in some activities.
- Must have ability to grasp complex problems and develop solutions, and possess in-depth background in high-performance computing.
- Excellent interpersonal, communications, presentation, and customer service skills and must have an outstanding track record of meeting customer expectations.
- Have demonstrated ability to clearly communicate ideas or technical concepts to a wide audience, both written and orally and experience developing and maintaining documentation.
- Required to work standard Lab working hours. Off-hours and weekend work will be required to support systems and users.
Desired:
- Experience implementing banking and allocation configurations for large systems.
- System administration experience, and experience implementing HPC storage subsystems (GPFS, Lustre, Panasas).
- Parallel programming for scientific applications experience.
- Programming experience with scientific software packages such as Matlab or IDL.
NOTES:
This is a two-year (Term) full time appointment, with the possibility of renewal and with the possibility of conversion to career based upon satisfactory job performance, continuing availability of funds, and ongoing operational needs.
Equal Employment Opportunity
Berkeley Lab is an affirmative action/equal opportunity employer committed to the development of a diverse workforce.
23685
|
