TY - JOUR
T1 - Stampede
T2 - A Cluster Programming Middleware for Interactive Stream-Oriented Applications
AU - Ramachandran, Umakishore
AU - Nikhil, Rishiyur S.
AU - Rehg, James M.
AU - Angelov, Yavor
AU - Paul, Arnab
AU - Adhikari, Sameer
AU - Mackenzie, Kenneth M.
AU - Harel, Nissim
AU - Knobe, Kathleen
N1 - Funding Information:
degree in computer science from the University of Wisconsin, Madison, in 1986, and is currently a professor in the College of Computing at the Georgia Institute of Technology. His fields of interest include parallel and distributed systems, computer architecture, and operating systems. Currently, he is leading a US National Science Foundation-ITR funded project that investigates the programming idioms and runtime systems for a distributed sensing infrastructure. He is the recipient of US National Science Foundation PYI Award in 1990, the Georgia Tech doctoral thesis advisor award in 1993, and the College of Computing Outstanding Senior Research Faculty award in 1996.
Funding Information:
A number of people have contributed to the Stampede project. Bert Halstead, Chris Jeorg, Leonidas Kontothanas-sis, and Jamey Hicks contributed during the early stages of the project. Dave Panariti developed a version of CLF for Alpha Tru64 and Windows. Members of the “ubiquitous presence” group at Georgia Tech continue to contribute to the project: Bikash Agarwalla, Matt Wolenetz, Hasnain Mandviwala, Phil Hutto, Durga Devi Mannaru, Namgeun Jeong, and Ansley Post deserve special mention. Russ Keldorph and Anand Lakshminarayanan developed an audio and video meeting application. Irfan Essa and Arno Schoedl provided the background necessary for understanding the video texture application. This work was done while R.S. Nikhil was at Compaq CRL. This work has been funded in part by a US National Science Foundation ITR grant CCR-01-21638, US National Science Foundation grant CCR-99-72216, HP/Compaq Cambridge Research Lab, the Yamacraw project of the State of Georgia, and the Georgia Tech Broadband Institute. The equipment used in the experimental studies is funded in part by a US National Science Foundation Research Infrastructure award EIA-99-72872, and Intel Corp.
PY - 2003/11
Y1 - 2003/11
N2 - Emerging application domains such as interactive vision, animation, and multimedia collaboration display dynamic scalable parallelism and high-computational requirements, making them good candidates for executing on parallel architectures such as SMPs and clusters of SMPs. Stampede is a programming system that has many of the needed functionalities such as high-level data sharing, dynamic cluster-wide threads and their synchronization, support for task and data parallelism, handling of time-sequenced data items, and automatic buffer management. In this paper, we present an overview of Stampede, the primary data abstractions, the algorithmic basis of garbage collection, and the issues in implementing these abstractions on a cluster of SMPS. We also present a set of micromeasurements along with two multimedia applications implemented on top of Stampede, through which we demonstrate the low overhead of this runtime and that it is suitable for the streaming multimedia applications.
AB - Emerging application domains such as interactive vision, animation, and multimedia collaboration display dynamic scalable parallelism and high-computational requirements, making them good candidates for executing on parallel architectures such as SMPs and clusters of SMPs. Stampede is a programming system that has many of the needed functionalities such as high-level data sharing, dynamic cluster-wide threads and their synchronization, support for task and data parallelism, handling of time-sequenced data items, and automatic buffer management. In this paper, we present an overview of Stampede, the primary data abstractions, the algorithmic basis of garbage collection, and the issues in implementing these abstractions on a cluster of SMPS. We also present a set of micromeasurements along with two multimedia applications implemented on top of Stampede, through which we demonstrate the low overhead of this runtime and that it is suitable for the streaming multimedia applications.
KW - Cluster computing
KW - Garbage collection
KW - Middleware
KW - Streaming applications
KW - Virtual time
UR - http://www.scopus.com/inward/record.url?scp=0346896344&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0346896344&partnerID=8YFLogxK
U2 - 10.1109/TPDS.2003.1247674
DO - 10.1109/TPDS.2003.1247674
M3 - Article
AN - SCOPUS:0346896344
SN - 1045-9219
VL - 14
SP - 1140
EP - 1154
JO - IEEE Transactions on Parallel and Distributed Systems
JF - IEEE Transactions on Parallel and Distributed Systems
IS - 11
ER -