davidepatti
commented
6 years ago Each simulation uses a different random seed, so that you can get statistically significant results by using the average of multiple simulations. This is necessary, since some feature are only probabilistically determined , for example, the packet injection rates. If you need to repeat a specific simulation, use -seed to force a specific random seed.
dear prof when i repeat simulation for the same configuration i receive different results for the average delay and max delay is this normal ??
example of the output: % Total received packets: 6855 % Total received flits: 13712 % Received/Ideal flits Ratio: 0.952222 % Average wireless utilization: 0 % Global average delay (cycles): 8.39344 % Max delay (cycles): 37 % Network throughput (flits/cycle): 1.52356 % Average IP throughput (flits/cycle/IP): 0.0952222 % Total energy (J): 6.86278e-06 % Dynamic energy (J): 1.08195e-07 % Static energy (J): 6.75459e-06
second simulation output % Total received packets: 6990 % Total received flits: 13980 % Received/Ideal flits Ratio: 0.970833 % Average wireless utilization: 0 % Global average delay (cycles): 8.40544 % Max delay (cycles): 35 % Network throughput (flits/cycle): 1.55333 % Average IP throughput (flits/cycle/IP): 0.0970833 % Total energy (J): 6.86384e-06 % Dynamic energy (J): 1.09253e-07 % Static energy (J): 6.75459e-06
and so on whenever i restart simulation for the same configuration