Check out the new USENIX Web site. next up previous
Next: About this document ... Up: Maximizing Beowulf Performance Previous: Availability

Bibliography

beowulf
See https://www.beowulf.org and links thereupon for a full description of the beowulf project, access to the beowulf mailing list, and more.

Amdahl
Amdahl's law was first formulated by Gene Amdahl (working for IBM at the time) in 1967. It (and many other details of interest to a beowulf designer or parallel program designer) is discussed in detail in the following three works, among many others.

Amalsi
G. S. Amalsi and A. Gottlieb, Highly Parallel Computing (2nd edition), Benjamin/Cummings, 1994.

Foster
I. Foster, Designing and Building Parallel Programs, Addison-Wesley, 1995. Also see the online version of the book at Argonne National Labs, https://www-unix.mcs.anl.gov/dbpp/

Kumar
V. Kumar, A. Grama, A. Gupta, and G. Karypis, Introduction to Parallel Computing, Design and Analysis of Algorithms, Benjamin/Cummings, 1994.

lmbench
A microbenchmark toolset developed by Larry McVoy and Carl Staelin of Bitmover, Incorporated. GPL plus special restrictions. See https://www.bitmover.com/lmbench/.

netperf
A network performance microbenchmark suite developed under the auspices of the Hewlett-Packard company. It was written by a number of people, starting with Rick Jones. Non-GPL open source license. See https://www.netperf.org/.

cpu-rate
A crude tool for measuring ``bogomflops'' written by Robert G. Brown and adapted for this paper. GPL. See https://www.phy.duke.edu/brahma.

Eden
The ``Eden'' beowulf consists of lucifer, abel, adam, eve, and sometimes caine and lilith. It lives in my home office and is used for prototyping and development.

profiling
Robert G. Brown, The Beowulf Design: COTS Parallel Clusters and Supercomputers, tutorial presented for the Extreme Linux Track at the 1999 Linux Expo in Raleigh, NC. Linked to https://www.phy.duke.edu/brahma, along with several other introductory papers and tools of interest to beowulf developers.

ATLAS
Automatically Tuned Linear Algebra Systems, developed by Jack Dongarra, et. al. at the Innovative Computing Laboratory of the University of Tennessee. Non-GPL open source license. See https://www.netlib.org/atlas.



Robert G. Brown 2000-08-28