The group’s mission is to research, design, and implement new models, features and optimizations of parallel programming languages and their execution environments for leading-edge High Performance Computing (HPC) systems. Research topics include but are not limited to design, implementation and optimization of parallel programming languages, compilers, operating systems, and advanced communication and run-time libraries. This builds on the group’s long history of leadership in this area with the development of the GASNet communication library, UPC programming language, UPC++ library, data-centric programming models, Operating Systems / Runtimes, and the software stack for quantum computers.
What You Will Do:
- Develop software for research related to the field of Partitioned Global Address Space programming models and their applications in HPC computing.
- Analyze the communication performance and inter-node scalability characteristics of applications and middleware which are candidates for use of UPC++ on present and future HPC systems.
- Develop and deploy algorithms using UPC++ to mitigate on-node or inter-node communication or other scalability bottlenecks in these applications and middleware.
- Participate in a multidisciplinary team involving computer scientists, engineers, mathematicians, and domain scientists with the goals of (a) developing and deploying UPC++ for the solution of scientific and engineering problems and (b) improving UPC++ and GASNet-EX to better meet the needs of their users.
- Document work and results in the form of technical reports, journal papers and conference proceedings.
- Publish work and present results at scientific meetings.
What is Required:
- Bachelor’s degree in Computer Science or a related field and a minimum of 8 years of related experience or equivalent experience.
- Experience implementing high-quality software in the field of High Performance Computing (HPC).
- Experience with design and implementation of applications using, or middleware implementing, the Partitioned Global Address Space (PGAS) model.
- Parallel programming in HPC applications and/or middleware using MPI, CUDA or a PGAS model.
- Proven consensus builder in a highly collaborative environment.
- Excellent written and verbal communication skills.
- Experience in parallel/distributed algorithm design and implementation, especially with respect to the use of non-blocking communications and asynchronous execution.
- Established track record of research accomplishments published in the peer-reviewed literature.
The posting shall remain open until the position is filled.
- This is a full-time 2 year, term appointment with the possibility of extension or conversion to Career appointment 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.
- Relocation for this position is optional, but the applicant must physically reside in the United States. The employee will be expected to attend regular virtual meetings with team and department members that are scheduled during business hours for US Pacific time.