Obtaining different results while executing a kernel on GPU or FPGA

Thank you for your answer!

The output is somehow very different. As example look at the following outputs obtained by GPU and FPGA:

FPGA:

w: 0

m: 0.1 51410 0.1 0

P: 8.25 13.5 0 0

-879.562 -1435.38 -0.0762125 -0.124711

0 0 8.25 13.5

0 0 13.5 26

w: 0

m: 0.1 56227.3 0.1 0

P: 8.25 13.5 0 0

-879.562 -1435.38 -0.0762125 -0.124711

0 0 8.25 13.5

0 0 13.5 26

w: 0

m: 0.1 95197.5 0.1 0

P: 8.25 13.5 0 0

-879.562 -1435.38 -0.0762125 -0.124711

0 0 8.25 13.5

0 0 13.5 26

w: 0

m: 0.1 85130.8 0.1 0

P: 8.25 13.5 0 0

-879.562 -1435.38 -0.0762125 -0.124711

0 0 8.25 13.5

0 0 13.5 26

w: 0

m: 0.1 68243.1 0.1 0

P: 8.25 13.5 0 0

-879.562 -1435.38 -0.0762125 -0.124711

0 0 8.25 13.5

0 0 13.5 26

w: 0

m: 0 0 0 0

P: 0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

GPU:

w: 0

m: 4.48707 0 2.19607 0

P: 8.17379 13.3753 0 0

13.5 26 0 0

0 0 8.17379 13.3753

0 0 13.5 26

w: 0

m: 4.89815 0 2.33002 0

P: 8.17379 13.3753 0 0

13.5 26 0 0

0 0 8.17379 13.3753

0 0 13.5 26

w: 0

m: 8.22379 0 2.51016 0

P: 8.17379 13.3753 0 0

13.5 26 0 0

0 0 8.17379 13.3753

0 0 13.5 26

w: 0

m: 7.36467 0 3.0552 0

P: 8.17379 13.3753 0 0

13.5 26 0 0

0 0 8.17379 13.3753

0 0 13.5 26

w: 0

m: 5.92356 0 2.4224 0

P: 8.17379 13.3753 0 0

13.5 26 0 0

0 0 8.17379 13.3753

0 0 13.5 26

w: 0

m: 0 0 0 0

P: 0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0