Answered step by step
Verified Expert Solution
Question
1 Approved Answer
Problem 1: We discussed in class the column sweep algorithm as it applies to parallelize the following matrix recurrence problem: X = C + AX
Problem 1: We discussed in class the column sweep algorithm as it applies to parallelize the following matrix recurrence problem: X = C + AX C = (G, ,Cn)T X = (X1 , , xn)T where A was a n n lower-triangular matrix, C is a column vector and X is a variable column vector. In this problem, you will apply the column sweep algorithm to parallelize the same type of recurrence but matrix A will be a general n n matrix. C1 C2 C3 11 a12 a13 a14 121 22 23 124 31 a32 33 434 41 42 43 44 We will use upto n processors operating in SIMD mode (single-instruction/multiple-data) a) Describe the major steps of the column sweep algorithm as they apply to this problem. Show the memory data structuring of C, A and X that supports this SIMD computation, assuming every processor has its own memory module b) What is the speedup and efficiency of the parallel scheme in comparison to a single processor solution How about the utilization? c) By careful consideration of the workload, it is possible to reduce the resource requirements by al- locating non-overlapping (in time) computation steps into the same processor. The overall parallel time should remain the same. Describe your processor allocation scheme. How many processors do you need at most. d) Suppose there are only m processors and n memory modules, respectively, where m
Step by Step Solution
There are 3 Steps involved in it
Step: 1
Get Instant Access to Expert-Tailored Solutions
See step-by-step solutions with expert insights and AI powered tools for academic success
Step: 2
Step: 3
Ace Your Homework with AI
Get the answers you need in no time with our AI-driven, step-by-step assistance
Get Started