Sidebar
This is an old revision of the document!
Table of Contents
SIG BUILD
Officers
- Student Adviser - Professor M. Theys
- President - Robert Blackwood
- Vice President - Andrew Sennebogen
- Systems Administrator - Walter Dworak
Proposals
First revision of proposals
Final order list (submitted to CS dept)
Hardware 101
Hardware Guides
Inventory
Projects
VRUPL Lab
PROJECT LEADER: Robert Blackwood
DESCRIPTION: Dr. Bell's VRUPL lab is in need of repair and networking. We are going to assist in getting the lab into working condition.
Status
In progress.
COMPLETED: New AD Server aka "dvorak"
PROJECT LEADER: Walter Dworak
DESCRIPTION: Moving the DCHP and ADDC roles, and other services to a new Windows Server Machine. The old machine is running Windows Server 2008 and the new machine is running Windows Server 2008 R2, so the change should not be as hard as moving from an NT4or5 to NT6 or moving from an NT based system to linux.
PURPOSE: The current machine charged with these roles, amadeus, is not meeting the necessary performance requirements to be a reliable sever, so the jobs are being moved to a new machine.
FURTHER READING: Any Microsoft documents related to moving Server roles between NT6 systems, and any other relevant documents from other sources.
Status
All tasks are completed.
P4 Cluster
PROJECT LEADER: Andrew Sennebogen
DESCRIPTION: ACCC P4 machines have been networked together in a commodity (Beowulf) computing cluster.
PURPOSE: The experience of building, maintaining, and utilizing a working computing cluster - including (but not limited to): netbooting, networking, MPI, parallelization of algorithms, UNIX, etc.
FURTHER READING: Computer Cluster (http://en.wikipedia.org/wiki/Cluster_%28computing%29)
Status
Andrew has (how many?) PC's networked together, and is working on getting different system software installed in all of them.
CUDA Machine
PROJECT LEADER: Jeremy Ervin
DESCRIPTION: The CUDA machine is a PC built with 4 high-end NVIDIA graphics cards installed. The goal is to harness tremendous computational power of these GPUs to do parallel computing.
PURPOSE: This project serves as a platform for computer science students at UIC to learn about parallel programming and GPU computing.
FURTHER READING: Compute Unified Device Architecture (http://en.wikipedia.org/wiki/CUDA)
Status
Jeremy and Robert have gotten the CUDA machine up and running with 3.5 cards working in Windows 7. All of the NVIDIA device drivers and updates have been installed, and the CUDA packages as well (including samples).
Supercomputer
PROJECT LEADER: Josh Herman
DESCRIPTION: Two supercomputers were donated to ACM member Josh Herman by professor Jan Verschelde. Josh is working to get these clustered and up and running as part of a research project.
PURPOSE: Josh Herman - The Chicago-01 supercomputer project is a research oriented supercomputer. Its primary goals are to investigate the following: Run code such as PHCPack (the implementation of polynomial homotopies exploiting structure in order to better approximate all isolated solutions) or basically find an optimal solution for hard polynominal systems. Another goal is to perform research on a novel network architecture utilizing artin braid groups as a algebra of networks. This will allow for the addition and subtraction of nodes in the network to be topologically invariant by the use of concatenation of the artin braid group. The last goal is to explore the use of GPU and possibly reconfigurable computing architectures like FPGA gates for the optimization of tasks on the supercomputer.
Suggestion: Edit Purpose in such a way that people can understand what you are talking about
Schedules
Robert
