Undergraduate: B.S. in Computer Science, Universitatea "Politehnica"
Bucuresti, Bucharest, Romania, June 1995
Graduate: M.S. in Computer Science, Universitatea "Politehnica"
Bucuresti, Bucharest, Romania, June 1996
Graduate: M.S. in Computer Science, University of Chicago, August
1999
Fifth year of University of Chicago's Computer Science Ph.D. program,
October 1997 - present. Advisor: Dr. Ian Foster.
Adriana Iamnitchi
Graduate Student - Department of Computer Science
University of Chicago
My current research plan is to look at scalable middleware services
for next-generation network applications, focusing on resource discovery.
While the number of Internet-based applications, whether information
providers (like web servers) or specialized computations (problem solvers
like NetSolve and NEOS) is growing fast, a new problem arises: how to
locate resources (computers, storage systems, applications, instruments,
etc.). A resource discovery mechanism should help users find resources
in concordance with a policy specified independently by the user (as
a request) and by the service provider (as an offer). The main challenges
in designing a resource discovery mechanism are scalability (because
of the difficult to predict, large number of users and resources) and
heterogeneity (various resources, with different owners and therefore
different policies).
My goal is to explore the design alternatives of a very scalable resource
discovery protocol, suitable for millions of heterogeneous entities,
with various policies and needs.
On Fully Decentralized Resource Discovery in Grid Environments.
Adriana Iamnitchi and Ian Foster. International Workshop on Grid Computing,
Denver, November 2001. http://people.cs.uchicago.edu/~anda/papers/GC2001.ps
Performance Predictions for a Numerical Relativity Package in Grid
Environments. Matei Ripeanu, Adriana Iamnitchi, Ian Foster. To appear
in International Journal on High Performance Computing Applications,
Volume 15, Number 4, November 2001. http://people.cs.uchicago.edu/~anda/papers/cactus_perf_journal.ps