A Game-Theoretic Approach for Run-Time Distributed Optimization on MP-SoC

Table 1

Convergence speeds.

	Average^*		Worst^*
Platform size	Game cycles	Time [𝜇s]	Game cycles	Time [𝜇s]
9 PEs	5	4.10	17	12.10
16 PEs	8	6.51	23	18.72
25 PEs	10	9.00	28	22.90
36 PEs	13	10.69	32	26.30
49 PEs	15	12.39	34	28.08
64 PEs	16	13.28	36	29.88
81 PEs	18	15.01	37	30.86
100 PEs	20	16.76	40	33.52

^*From probabilistic analysis.