DREU Summer 2010                                        Andres Fernandez

Home  Week 1  Week 2  Week 3   Week 4  Week 5  Week 6  Week 7  Week 8   Final Report     Week 5 Report June 28 - July 2 2010  This week my advisor and I decided to make a change to the memory storage that is holding the vector. The change is going to be a dual FIFO. The first FIFO is going to be loaded up with the data and after it is full the second FIFO is going to be loaded. While the second FIFO is loading up the first FIFO is going to use the data stored for the vector matrix multiplication. Each of the FIFO's has its own feedback so it can do the vector matrix multiplication. This set-up is good for a dense matrix multiplication, but for a structured matrix (lower and upper triangle for example) we are not sure whether there will be a bottle neck. You can reach me at fernandez.andres@sbcglobal.net