- Parallel Computing (Dr. Jones, 20090527-20090529)
is a free Firefox extension for collecting, managing, and citing your research sources. John Ringland
April 25, 2009, at 10:42 AM
Random number generator speeds
Please test your own random number generator for speed and enter the results below.
For the "RNs per second per GHz", divide the number of RNs generated by the
real time in seconds (on a machine where nothing much else is running) and by the speed of your CPU in GHz".
(On Linux, type "cat /proc/cpuinfo" to get CPU info.)
|Nominal distribution||Generator name||RNs per second per GHz||Test computer||Tester||Date|
|uniform in (0,1)||uni() (C++ Boost)||47 million||IBM laptop 1.7GHz, 1.25GB RAM||Chen Zhang||2/5/09|
|uniform in (0,1)||drand48 (C++ math.h)||15 million||IBM laptop 1.7GHz, 1.25GB RAM||Chen Zhang||2/5/09|
|uniform in [0,1)||Maple RandomTools[MersenneTwister] GenerateFloat64()||37 thousand||Intel P4 2.93GHz, 2.5GB RAM||John Ringland||2/5/09|
|uniform in (0,1)||rng.txt (Maple simple LCG )||16 thousand||Intel P4 2.93GHz, 2.5GB RAM||John Ringland||2/5/09|
|uniform in (0,1)||Octave rand()||160 thousand||Intel Core 2 Duo 1.6GHz||Steven Leuthe||3/13/09|
|uniform in (0,1)||Matlab rand()|| || || || |
|binomial(n=200,p=0.5)||numerical recipes(using uni() from Boost)||0.2 million||IBM laptop 1.7GHz, 1.25GB RAM||Chen Zhang||2/5/09|
|poisson(mean=200.0)||numerical recipes(using uni() from Boost)||0.5 million||IBM laptop 1.7GHz, 1.25GB RAM||Chen Zhang||2/5/09|
|uniform in (0,1)*||Maple Statistics[Sample]()||14 million||Intel P4 2.40GHz, 2.0GB RAM||Matt Westley||2/9/09|
|uniform in (0,1)||Mathematica, Random||125 thousand||AMD Athlon Dual Core 3800+ 2.0GHz, 3.0GB RAM, Windows Vista||Benjamin Kanouse||2/10/09|
*but it can generate from a lot of different distributions
Work in progress
Latex Mathematical Symbols
Tarball used in Feb 9 meeting
SYLLABUS brainstorming page
NOTE: The list of topics below should be related whenever possible to the current research projects, and labeled by the project mentors (Bayram: Ba, Biondini: Bi, Ringland: R, Sen: S). These topics will be broken up into one hour modules.
List of topics of interest followed (whenever possible) by the project it is needed for:
- Operating system, Networking and hardware
- Ubuntu OS and open source software: GUI vs command line, file paths, permissions, file management, ssh, scp, OpenOffice, ...
- Basic networking concepts (enough for students to understand how to troubleshoot communication problems). For example: IP and MAC addressing; the concept of digital communications and packets; ports on the computers; security with IP tables as implemented with Firestarter;
- Basic facts about the hardware used in Computational Science: parts of a computer and how they affect the speed of computations;
- Group work and communication software
- Web pages, Wikis, Skype, ...
- Scientific typesetting and illustration
- Latex: Should we use Latex with a front end like Kile? Lyx?
- Vector graphics editing: Adobe Illustrator via VirtualBox/Windows
- Programming software: Matlab, Mathematica, Maple, C++.
- Parallel computing
- NEW TO ME: parallel computing using graphic cards (GPU computing).
- On-line and in person research skills
- Journals, books, and other sources
- Bibliographic management: JabRef, bibtex
- How to evaluate Journals and Conferences (impact factor, publishing houses, etc.)