I am using the AD9915 evaluation board with SYSCLK = 2.5 GHz and I am finding spurious content that is not predicted by the ADIsimDDS tool. Two specific examples are for fOUT set at 927.5 MHz there is a spur at 947.5 MHz at -71 dBc and for fOUT set at 932.5 MHz there is a spur at 942.5 MHz at -70 dBc. There are many other instances such as these and they are well below the cut-off of my reconstruction filter. Are these spurs expected to be there or is there something wrong with my evaluation board or how I'm using it? Can anything be done to reduce or eliminate these spurs? Any help in this matter is greatly appreciated.