Question 1

Read Inada and McCool, "Compressed Lossless Texture Representation and Caching", In Proceedings of GH 2006: The ACM SIGGRAPH/Eurographics Symposium on Graphics Hardware

  1. List the correspondences between concepts in this paper and CPU cache concepts covered in this class.
  2. Using CPU numbers (miss rates, etc.) in the book, from a paper, or that you find online, evaluate the speedup or slowdown of using compressed data in this way in a CPU cache. Show your work, but you should produce a numerical result. For each number that goes into the computation, clearly indicate what it is and where it comes from (i.e. book, paper or other source; number in the text, table or graph; if from a graph, which line of the graph and what you estimated the value to be, etc.).

Question 2

Read Younis, Bengad and Akkaya, "Base-Station Repositioning for Optimized Performance of Sensor Networks," In Proceedings of VTC 2003-Fall: the 58th IEEE Vehicular Technology Conference, October 6-9, 2003, pp. 2956-2960.

  1. What performance measure(s) are used to evaluate the success of the repositioning strategy?
  2. What performance measures(s) are used to decide on the repositioning?
  3. Why are these not the same?
  4. What is the approximate "speedup" of each term graphed in the results section?
  5. The "workload" in this case is the random distribution of sensor nodes. What other workloads that might make sense? How might the results in these cases.