Double Product combine1: Maximum use of data abstraction:
11.50 cycles/element
Double Product combine2: Take vec_length() out of loop:
11.50 cycles/element
Double Product combine3: Array reference to vector data:
4.79 cycles/element
Double Product combine4: Array reference, accumulate in temporary:
4.76 cycles/element
Double Product combine4p: Pointer reference, accumulate in temporary:
4.76 cycles/element
Double Product Array code, unrolled by 2:
4.76 cycles/element
Double Product combine5p: Pointer code, unrolled by 3, for loop:
4.76 cycles/element
Double Product Array code, unrolled by 3, while loop:
4.76 cycles/element
Double Product Array code, unrolled by 4:
4.76 cycles/element
Double Product Array code, unrolled by 8:
4.77 cycles/element
Double Product Array code, unrolled by 16:
4.77 cycles/element
Double Product Pointer code, unrolled by 2:
4.77 cycles/element
Double Product Pointer code, unrolled by 3:
4.76 cycles/element
Double Product Pointer code, unrolled by 4:
4.76 cycles/element
Double Product Pointer code, unrolled by 8:
4.77 cycles/element
Double Product Pointer code, unrolled by 16:
4.76 cycles/element
Double Product combine6: Array code, unrolled by 2, Superscalar x2:
2.38 cycles/element
Double Product Array code, unrolled by 4, Superscalar x2:
2.39 cycles/element
Double Product Array code, unrolled by 8, Superscalar x2:
2.41 cycles/element
Double Product Array code, unrolled by 3, Superscalar x3:
1.59 cycles/element
Double Product Array code, unrolled by 4, Superscalar x4:
1.19 cycles/element
Double Product Array code, unrolled by 8, Superscalar x4:
1.23 cycles/element
Double Product Array code, unrolled by 6, Superscalar x6:
0.98 cycles/element
Double Product Array code, unrolled by 8, Superscalar x8:
1.00 cycles/element
Double Product Array code, unrolled by 10, Superscalar x10:
0.99 cycles/element
Double Product Array code, unrolled by 12, Superscalar x6:
0.98 cycles/element
Double Product Array code, unrolled by 12, Superscalar x12:
0.99 cycles/element
Double Product Pointer code, unrolled by 8, Superscalar x2:
2.40 cycles/element
Double Product Pointer code, unrolled by 8, Superscalar x4:
1.23 cycles/element
Double Product Pointer code, unrolled by 8, Superscalar x8:
0.99 cycles/element
Double Product Pointer code, unrolled by 9, Superscalar x3:
1.64 cycles/element
Double Product Array code, Unroll x2, Superscalar x2, noninterleaved:
2.39 cycles/element
Double Product Array code, unrolled by 2, different associativity:
2.39 cycles/element
Double Product Array code, unrolled by 3, Different Associativity:
1.59 cycles/element
Double Product Array code, unrolled by 4, Different Associativity:
1.19 cycles/element
Double Product Array code, unrolled by 6, Different Associativity:
0.98 cycles/element
Double Product Array code, unrolled by 8, Different Associativity:
0.99 cycles/element
Double Product SSE code, 1*VSIZE-way parallelism:
2.39 cycles/element
Double Product SSE code, 2*VSIZE-way parallelism:
1.19 cycles/element
Double Product SSE code, 4*VSIZE-way parallelism:
0.64 cycles/element
Double Product SSE code, 8*VSIZE-way parallelism:
0.48 cycles/element
Double Product SSE code, 12*VSIZE-way parallelism:
0.48 cycles/element
Double Product SSE code, 2*VSIZE-way parallelism, reassociate:
1.20 cycles/element
Double Product SSE code, 4*VSIZE-way parallelism, reassociate:
0.63 cycles/element
Double Product SSE code, 8*VSIZE-way parallelism, reassociate:
0.49 cycles/element