|
Contact
information
|
Email: rob dot schreiber at hp dot com
|
Phone: (650) 857-8156
Physical Address: 1501 Page Mill Road, MS 1177, Palo Alto, CA 94304
Bio
Rob
Schreiber is Assistant Director of the Exascale Computing Lab at Hewlett
Packard Laboratories. He is known for basic research in sequential and
parallel algorithms for matrix computation, and compiler optimization for
parallel languages. Rob has been a professor of Computer Science at
Stanford University and at RPI, was Chief Scientist at Saxpy Computer, and
a research scientist at the NASA Ames Research Center. He was a developer
of the sparse-matrix extension of Matlab, a leading designer of the High
Performance Fortran programming language, and one of the developers of the
NAS parallel benchmarks. He wrote the matrix computation libraries at
Maspar. At HP, Rob was a technical leader and an implementer of PICO, a
tool for hardware synthesis from high-level specifications. His current
research is in algorithms and architectures for high-performance computing
and data analysis.
|
·
Graph
Clustering:
With Bob Tarjan, Nina Mishra, Dennis Wilkinson, and Bei Wang, I've
developed new approaches to finding clusters of graph vertices that have
more connections to one another than would be expected given their
degree.
·
Role
Discovery:
Discovering the roles or job functions in an organization from acess
control lists can be a useful first step in introducing role-based access
control.
·
Nanophotonic
Architecture:
The microprocessor of the future will consist of many CPU
cores. What isn't clear is how they will talk to each other
and to memory. Our project looks at the use of integrated photonic
communication devices to raise bandwidths and reduce the energy cost of
communication.
·
The
Parallel Programming Features of Matlab: I've worked with Mathworks on
the new features of Matlab that allow a programmer to write code for a
parallel cluster.
·
Other
topics in Parallel Programming: Recent work on synchronization cost,
on memory management.
·
Some work on computer vision: There are nice
applications of linear algebra, theoretical and computational.
|
Slides
from some recent talks
Publications
A list of my selected paper publications is here.
Data
The user-permission relations (bipartite graphs) used for the experiments
in our SACMAT 08 paper, in a zip file: sacmat_relations.zip.
The
covers generated by the greedy heuristic with lattice-based postprocessing
are in another: lattice_covers.zip
The
exact covers (minimum biclique covers) generated by our problem reduction /
graph coloring approach are in a third: exact_covers.zip
|
|
|
