Float Sum combine1: Maximum use of data abstraction:
27.97 cycles/element
Float Sum combine2: Take vec_length() out of loop:
26.04 cycles/element
Float Sum combine3: Array reference to vector data:
5.00 cycles/element
Float Sum combine4: Array reference, accumulate in temporary:
5.00 cycles/element
Float Sum combine4p: Pointer reference, accumulate in temporary:
5.00 cycles/element
Float Sum Array code, unrolled by 2:
5.00 cycles/element
Float Sum combine5p: Pointer code, unrolled by 3, for loop:
5.00 cycles/element
Float Sum Array code, unrolled by 3, while loop:
5.01 cycles/element
Float Sum Array code, unrolled by 4:
5.00 cycles/element
Float Sum Array code, unrolled by 8:
5.01 cycles/element
Float Sum Array code, unrolled by 16:
5.00 cycles/element
Float Sum Pointer code, unrolled by 2:
5.01 cycles/element
Float Sum Pointer code, unrolled by 3:
4.99 cycles/element
Float Sum Pointer code, unrolled by 4:
5.00 cycles/element
Float Sum Pointer code, unrolled by 8:
4.99 cycles/element
Float Sum Pointer code, unrolled by 16:
5.00 cycles/element
Float Sum combine6: Array code, unrolled by 2, Superscalar x2:
2.50 cycles/element
Float Sum Array code, unrolled by 4, Superscalar x2:
2.51 cycles/element
Float Sum Array code, unrolled by 8, Superscalar x2:
2.50 cycles/element
Float Sum Array code, unrolled by 3, Superscalar x3:
2.00 cycles/element
Float Sum Array code, unrolled by 4, Superscalar x4:
2.01 cycles/element
Float Sum Array code, unrolled by 8, Superscalar x4:
2.00 cycles/element
Float Sum Array code, unrolled by 6, Superscalar x6:
2.01 cycles/element
Float Sum Array code, unrolled by 8, Superscalar x8:
2.01 cycles/element
Float Sum Array code, unrolled by 10, Superscalar x10:
2.00 cycles/element
Float Sum Array code, unrolled by 12, Superscalar x6:
2.00 cycles/element
Float Sum Array code, unrolled by 12, Superscalar x12:
2.00 cycles/element
Float Sum Pointer code, unrolled by 8, Superscalar x2:
2.51 cycles/element
Float Sum Pointer code, unrolled by 8, Superscalar x4:
2.00 cycles/element
Float Sum Pointer code, unrolled by 8, Superscalar x8:
1.98 cycles/element
Float Sum Pointer code, unrolled by 9, Superscalar x3:
2.01 cycles/element
Float Sum Array code, Unroll x2, Superscalar x2, noninterleaved:
2.50 cycles/element
Float Sum Array code, unrolled by 2, different associativity:
2.75 cycles/element
Float Sum Array code, unrolled by 3, Different Associativity:
2.05 cycles/element
Float Sum Array code, unrolled by 4, Different Associativity:
2.01 cycles/element
Float Sum Array code, unrolled by 6, Different Associativity:
2.00 cycles/element
Float Sum Array code, unrolled by 8, Different Associativity:
2.00 cycles/element
Float Sum SSE code, 1*VSIZE-way parallelism:
1.25 cycles/element
Float Sum SSE code, 2*VSIZE-way parallelism:
0.82 cycles/element
Float Sum SSE code, 4*VSIZE-way parallelism:
0.50 cycles/element
Float Sum SSE code, 8*VSIZE-way parallelism:
0.58 cycles/element
Float Sum SSE code, 12*VSIZE-way parallelism:
0.44 cycles/element
Float Sum SSE code, 2*VSIZE-way parallelism, reassociate:
1.08 cycles/element
Float Sum SSE code, 4*VSIZE-way parallelism, reassociate:
0.89 cycles/element
Float Sum SSE code, 8*VSIZE-way parallelism, reassociate:
0.58 cycles/element