ST: CDA 6938 Multi-core/many-core architectures and programming




  1. Select an application that has rich data-level parallelism. Select your target hardware platform(s). You may also choose to perform cross-platform comparison. Sample topics: image compression/decompression; computer vision; ray tracing; encryption/decryption; etc. ( Here is a more detailed list of interesting applications extracted from the Matlab Image Processing toolbox)


  1. Write a 1-page proposal describing the application and the key algorithms. Report the group members (a group can have one or two students depending on the problem size) and the milestones of the project. (send your proposal via email to by 3/17/09)


  1. An in-class short presentation (5-10 minutes) of your project proposal on 3/19/09.


  1.  Final project presentation (~30 minutes). The content includes: the review of the application and the algorithms, the CPU implementation, GPGPU implementation, performance optimization, lesson learnt from the project, future work. (presentation date: TBA)


  1. Send your source code, a technical report, and your presentation slides to (due date: 4/30/09)