JACOBI:
  CUDA version.
  Apply Jacobi's iterative method to solve A*x=b
  System size is N = 10000

    I    B[i]      X[i]     Resid[I]

    0   1.28599         0  -1.28599
    1  -0.834521         0  0.834521
    2  0.373376         0  -0.373376
    3  -0.0978665         0  0.0978665
    4  0.827303         0  -0.827303
    5  -0.851767         0  0.851767
    6  -0.295335         0  0.295335
    7  0.923462         0  -0.923462
    8  -0.76665         0   0.76665
    9  0.352768         0  -0.352768

  RMS error = 0.70168

  Computation required 0.11 seconds

JACOBI:
  Normal end of execution.