Podcast
Questions and Answers
The number on an arrow indicates the time step during which the message was transferred.
The number on an arrow indicates the time step during which the message was transferred.
False (B)
In the matrix-vector multiplication example, each processor computes the product of the entire vector and matrix.
In the matrix-vector multiplication example, each processor computes the product of the entire vector and matrix.
False (B)
The final step of the matrix-vector multiplication involves a one-to-all broadcast of the vector.
The final step of the matrix-vector multiplication involves a one-to-all broadcast of the vector.
False (B)
The matrix is assigned to a processor grid with n x (n+1) processors.
The matrix is assigned to a processor grid with n x (n+1) processors.
The all-to-one reduction operations are performed using the multiplication operation.
The all-to-one reduction operations are performed using the multiplication operation.
The product vector is computed on the columns of processors.
The product vector is computed on the columns of processors.
The broadcast of the vector element is done concurrently for all n rows.
The broadcast of the vector element is done concurrently for all n rows.
Each processor computes the product of the entire matrix and vector.
Each processor computes the product of the entire matrix and vector.
The first step of the matrix-vector multiplication involves an all-to-one reduction operation.
The first step of the matrix-vector multiplication involves an all-to-one reduction operation.
The matrix is divided into submatrices and assigned to separate processors.
The matrix is divided into submatrices and assigned to separate processors.