76
edits
Changes
→-Amdahls Predicted speed increase
For this analysis - the Nvidia 1070 GPU is available for testing and has a total of 1920 CUDA cores.
S(n) = 1 / 1 - (0.6667) + 0.6667 / 1920 = 1 / 0.3333 + 0.00034723 = 1 / 0.333647239 = 2.99x
Therefore by Amdahl's law, there is a potential speed increase of 2.99x if this component were to be paralleled.
== -Additional Info Available @ ==
