CSCI 211 Group Project Outline

by

Marc Scott v.s. Aravind Tukakula v.s. Ian-Ning Young

Team :

Marc Scott

Aravind Tukakula

Ian-Ning Young

 

Topic: Java For High Performance Computing

Thesis: For Java offers a number of benefits as a language for High Performance Computing (HPC), it is important for us to know about what Java for HPC is, what the benefits Java for HPC are, and what the challenges and comments Java for HPC faces.

 

  1. Introduction to High-Performance Computing
    1. History/Background (by Marc Scott)
    2. The Java Grande Forum (by Ian-Ning Young)

     

  2. Architecture Issues
    1. Major Computer Systems (by Marc Scott)
    2. Platform (by Aravind Tukakula)
    3. Security (by Aravind Tukakula)
  3. Performance Issues
    1. Comparison with C/C++ (by Aravind Tukakula)
    2. Some General Performance Improvements left to make (by Aravind Tukakula)
    3. Performance Advantages/Disadvantages of current trends (by Aravind Tukakula)
  4. More High-Performance Java Issues
    1. Java Computing Frameworks and Collaborative environments (by Marc Scott)
    2. Java Compiling techniques (by Marc Scott)
    3. Java Tools for high Performance Computing (by Marc Scott)
    4. Performance Analysis (arrays, complex numbers, numerical calculations) (by Marc Scott)
  5. Parallel Computing
    1. vs. Serial Computing (by Marc Scott)
    2. Clusters (by Marc Scott)
    3. PVMs (by Aravind Tukakula)
    4. CORBA (by Aravind Tukakula)
    5. Multi-Threading (by Ian-Ning Young)
    6. API (by Ian-Ning Young)
    7. MPI (by Ian-Ning Young)
    8. MPJ (by Ian-Ning Young)
    9. RMI (by Ian-Ning Young)
  6. Numerical Issues
    1. Brief Overview of Numeric Analysis (by Aravind Tukakula)
    2. Floating-Point (by Ian-Ning Young)
    3. Complex Numbers (by Ian-Ning Young)
    4. Multi-Dimensional Arrays (by Ian-Ning Young)
    5. More Numeric Benchmarks (by Aravind Tukakula)
    6. JNL (Java Numerical Library) - Linear Algebra Classes (by Marc Scott)
  7. Summary of Challenges and Commends
    1. Past Challenges (by 3 group members)
    2. Current Trends (by 3 group members)
    3. Future Improvement (by 3 group members)

 

Conclusion: Java will eventually be the alternative for High Performance Computing after the performance issue improved/resolved.

 

 

 

1