EDUCATION:
  Princeton University, Ph.D. in Astrophysics, 1998
    Specialization: Computational Cosmology
    DoE Computational Science Graduate Fellow
  Pomona College, B.A. in Astrophysics

BIOGRAPHY:

  Dr. Kepner leads large scale computing research across MIT's largest laboratory (Lincoln Laboratory accounts for over half of all research performed at MIT).  Dr. Kepner is the most published author in the 60+ year history of Lincoln Laboratory.  His published works span signal processing, data mining, databases, high performance computing, graph algorithms, cyber security, visualization, cloud computing, random matrix theory, abstract algebra, bioinformatics, astronomy, physics, and astrophysics.  In addition he has authored two books on parallel computing and graph algorithms. He recently received Lincoln’s highest honor for technical excellence “For his leadership and vision in bringing supercomputing to Lincoln Laboratory through the establishment of LLGrid; his pivotal role in open systems for embedded computing; his creativity in developing a novel database management language and schema; and his contributions to the field of graph analytics.”  More recently, Dr. Kepner has been at the forefront of developing new signal processing technique for genetic sequence analysis and operating on data while it is stored in encrypted form.  Dr. Kepner is the Chair of the largest computing conference in New England (IEEE High Performance Extreme Computing) and Vice-Chair of SIAM Data Mining.  Dr. Kepner received his Ph.D. in Astrophysics from Princeton University in 1998.

  Jeremy is most proud of the opportunity he has had to be the principal architect, PI or otherwise co-lead several very talented teams.  These teams have produced a number of innovative technologies that have broken new ground in several domains and have thousands of users worldwide:

MatlabMPI
 - 1st widely usable MPI for MATLAB(R).
 - 1st MPI implemented using file IO.
pMatlab
 - 1st widely usable PGAS implementation in a high level language,
   and the basis of The Mathworks Parallel Computing Toolbox (PCT).
pMatlabXVM
 - 1st hierarchical PGAS library.
LLGrid
 - 1st 1000+ processor interactive/on-demand parallel computing system.
TX-2500
 - 1st petascale virtual memory system.
 - World's fastest Terabyte file sort (2006).
PVL (Parallel Vector Library)
 - 1st parallel object oriented signal processing libary.
VSIPL++ (Vector, Signal, and Image Processing Library)
 - 1st parallel object oriented open standard library.
PVTOL (Parallel Vector Tile Optimizing Libary)
 - 1st parallel/multi-core object oriented signal processing libary.
High Productivity Computing Systems
 - 1st system model of productivity for parallel computing systems.
HPC Challenge Benchmark
 - 1st parallel benchmark suite designed to fully stress the memory hierarchy.
 - 1st parallel benchmark suite designed to test programmability.
HPC Challenge Awards
 - 1st contest to recognize innovations in parallel programming.
HPEC Challenge Benchmark
 - 1st benchmark suite targeting HPEC systems.
HPEC Workshop
 - 1st workshop focused on High Performance Embedded Computing.
Dynamic Distributed Dimensional Data Model (D4M)
 - 1st associative algebra for graph signal processing.