Performance Engineering and Data Analytics
Argonne Leadership Computing Facility
Argonne National Laboratory
Stuart Building 204
Tuesday, April 24th, 2012
Dr. Kalyan Kumaran possesses a strong record of
accomplishments in both R&D and management. He has worked extensively in
optimizing and tuning numerical algorithms on parallel platforms. He
performed extensive hardware and profiling analysis of the SPEC
He was a key developer of the CPU benchmark suite, SPEC CPU2006, where he was project leader for 15 benchmark candidate codes. He also led the efforts in the development of SPEC MPI2007 and SPEC OMP2001, the first industry standard application benchmarks in MPI and OpenMP. Moreover, at Silicon Graphics, Kalyan managed a performance team that created several world records in industry-standard benchmarks, including top scores on the SGI Altix system at NASA Ames (52 TF) and STREAM Triad (1 TB/s).
He has held several leadership positions in performance engineering and for a decade assumed responsibility both as a team leader and manager of performance benchmarking projects.
Currently, as the ALCF Manager of Performance Engineering and Data Analytics, Kalyan is responsible for defining and building applications and performance engineering services for the ALCF. He leads an application services team in evaluating application performance, makes recommendations for algorithmic and implementation tuning strategies, and supervises the implementation of those recommendations.
Eleanor Taylor, 630/252-5510, email@example.com
Awards, Honors, and Memberships
SGI Systems Engineer of the Year, 2000.
SPEC – SPECtacular Contribution Awards, 2002, 2003, 2004, 2005.
Recognized by SGI’s Chief Technology Officer for contributions to engineering, Dec. 2003, Dec. 2004.
Member, SPEC Board of Directors, since 2003.
Chair, SPEC High-Performance Group, since 2003.
B.Tech (Honors), Indian Institute of Technology, 1988
M.S., Iowa State University, 1990
Ph.D., Iowa State University, 1994
High Performance Computing, Networking and data management
computational fluid dynamics
parallelizing scientific applications