cs8803SC_lecture22 - architecture Term Project Presentation...

Info iconThis preview shows pages 1–8. Sign up to view the full content.

View Full Document Right Arrow Icon
   CS8803SC Software and Hardware Cooperative Computing    AMD ATI/Brook+ Prof. Hyesoon Kim School of Computer Science Georgia Institute of Technology
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
  Brook/BrookGPU/Brook+ Brook+ AMD’s GPGPU language Based on Brook from Stanford BrookGPU: open source version of Brook for GPU computing High-level language, C extensions for the GPU
Background image of page 2
  Kernels/Streams Kernels are functions that run on the GPU. The kernel is invoked on every element of the stream. Streams 1D array: up to 8192 elements 2D array: up to 8192 x 8192 elements Virtualization is possible Streams cannot be accessed directly by the application: Data must be copied between streams and memory using streamRead and streamWrite Explicit read and write
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
  Brook+ http://coachk.cs.ucf.edu/courses/CDA6938/Broo Architecture http://coachk.cs.ucf.edu/courses/CDA6938/UC F_1_25_08.pdf
Background image of page 4
  Brook vs CUDA Both are based on C Brook is a little bit more high level CUDA is a little bit exposed to the
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 6
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 8
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: architecture Term Project Presentation Starts 9:00 AM and ends 11:15 AM on both Tuesday and Thursday 1 or 2 member tem: 13 min presentation + 2 min Q&A 3 member team 17 min presentation + 3 min Q&A Presentation Introduction Background Design overview Code snippets Implementation Performance optimization techniques Issues in speedup Verification Performance results Kernel execution time vs. application execution time Different blocks sizes, different thread sizes Data transfer time, computation time CUDA vs. OpenMP or MPI Extra: performance improvement vs. programming effort Conclusion & Future work What you have learned through the project Announcement Lab #3 due is coming. Next Monday (4/14) 11:55 pm Please submit a hardcopy of your report in the class (4/15) Check with the newsgroup...
View Full Document

This note was uploaded on 10/06/2010 for the course CS 8803 taught by Professor Staff during the Spring '08 term at Georgia Institute of Technology.

Page1 / 8

cs8803SC_lecture22 - architecture Term Project Presentation...

This preview shows document pages 1 - 8. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online