Various Possible Implementations
Broadcast
Y moves
W moves
Fan-In
Systolic
Bidirectional
Y in place
X in place
Unidirectional
Y in place
W in place
Running Example: Convolution
for (i=1; i ? n; i++)
for (j=1; j ? k; j++)
yi += wj * xi+j-1
Previous slide
Next slide
Back to first slide
View graphic version