t0 = 0.001 ms, t1 = 0.676 ms // non-unrolled, calls to sinf/cosf (SPU SIMD, power series, 4 terms)
t0 = 0.001 ms, t1 = 0.764 ms
t0 = 0.001 ms, t1 = 0.764 ms
t0 = 0.001 ms, t1 = 0.764 ms
t0 = 0.001 ms, t1 = 0.764 ms

t0 = 0.398 ms, t1 = 0.002 ms // unrolled x4, calls to sinf/cosf (SPU SIMD, power series, 4 terms)
t0 = 0.452 ms, t1 = 0.002 ms // 1.69 x
t0 = 0.452 ms, t1 = 0.002 ms
t0 = 0.451 ms, t1 = 0.002 ms
t0 = 0.452 ms, t1 = 0.002 ms

t0 = 0.474 ms, t1 = 0.002 ms // unrolled x4, calls to sinf/cosf (SPU SIMD, power series, 4 terms), no arith cooking
t0 = 0.474 ms, t1 = 0.002 ms
t0 = 0.420 ms, t1 = 0.002 ms
t0 = 0.474 ms, t1 = 0.002 ms
t0 = 0.475 ms, t1 = 0.002 ms

t0 = 0.200 ms, t1 = 0.002 ms // unrolled x4, custom vectorized sincos
t0 = 0.227 ms, t1 = 0.002 ms // 3.36 x
t0 = 0.227 ms, t1 = 0.002 ms
t0 = 0.227 ms, t1 = 0.002 ms
t0 = 0.227 ms, t1 = 0.002 ms

t0 = 0.221 ms, t1 = 0.002 ms // unrolled x4, custom vectorized sincos, no arith cooking
t0 = 0.249 ms, t1 = 0.002 ms
t0 = 0.249 ms, t1 = 0.002 ms
t0 = 0.249 ms, t1 = 0.002 ms
t0 = 0.249 ms, t1 = 0.002 ms