Problem was not enough memory in state array for thread/block size

    curandState * d_state;  
    cudaMalloc(&d_state, 195075 * sizeof(curandState) );

    __global__ void k_initRand(curandState *state, uint64_t seed){
        int tid = threadIdx.x + blockIdx.x * blockDim.x;
        curand_init(seed, tid, 0, &state[tid]);

    }

Out of bounds error was garbling the printf() data