Iteration 1
Iteration 3
IP - Sets of two cores
IP - Sets of single cores
IP - Sets of eight cores
IP - Sets of four cores
Iteration 2
Iteration 4
Single IP - Set of 16 cores
132
Network-on-Chip
Figure 5.5
Example binomial merging.
5.4.3 Hardware Cost Optimization
Since the routers are the dominant hardware components in an OCN, BMAP
algorithm attempts to optimize the same, using the traffic loading information generated at the previous stage. The routers, specially the buffers, consume a good amount of area and power in the network. This hardware cost
is optimized using the strategies as follows:
Iteration 3
IP - Sets of two cores
IP - Sets of single cores
IP - Sets of eight cores
IP - Sets of four cores
Iteration 2
Iteration 4
Single IP - Set of 16 cores
132
Network-on-Chip
Figure 5.5
Example binomial merging.
5.4.3 Hardware Cost Optimization
Since the routers are the dominant hardware components in an OCN, BMAP
algorithm attempts to optimize the same, using the traffic loading information generated at the previous stage. The routers, specially the buffers, consume a good amount of area and power in the network. This hardware cost
is optimized using the strategies as follows: