| Time | Implementation | Compiler | Benchmark date | SUPERCOP version |
|---|---|---|---|---|
| 26001 | moon/neon/32 | gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | 20161210 | 20161026 |
| 26036 | moon/neon/32 | gcc -mcpu=native -mfpu=neon-vfpv4 -Os | 20161210 | 20161026 |
| 27807 | moon/neon/32 | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | 20161210 | 20161026 |
| 32303 | moon/neon/32 | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | 20161210 | 20161026 |
| 32638 | moon/neon/32 | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | 20161210 | 20161026 |
| 33411 | moon/neon/32 | gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | 20161210 | 20161026 |
| 77859 | moon/armv6/32 | gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | 20161210 | 20161026 |
| 77884 | moon/armv6/32 | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | 20161210 | 20161026 |
| 78015 | moon/armv6/32 | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | 20161210 | 20161026 |
| 78532 | moon/armv6/32 | gcc -mcpu=native -mfpu=neon-vfpv4 -Os | 20161210 | 20161026 |
| 78565 | ref | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | 20161210 | 20161026 |
| 86880 | moon/armv6/32 | gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | 20161210 | 20161026 |
| 102890 | moon/armv6/32 | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | 20161210 | 20161026 |
| 109419 | ref | gcc -mcpu=native -mfpu=neon-vfpv4 -Os | 20161210 | 20161026 |
| 113105 | ref | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | 20161210 | 20161026 |
| 113710 | ref | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | 20161210 | 20161026 |
| 127500 | ref | gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | 20161210 | 20161026 |
| 131770 | ref | gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | 20161210 | 20161026 |
| 177109 | regs | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | 20161210 | 20161026 |
| 223462 | regs | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | 20161210 | 20161026 |
| 228669 | regs | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | 20161210 | 20161026 |
| 243002 | regs | gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | 20161210 | 20161026 |
| 248560 | regs | gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | 20161210 | 20161026 |
| 250703 | regs | gcc -mcpu=native -mfpu=neon-vfpv4 -Os | 20161210 | 20161026 |
| Compiler | Implementations |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/xop/32 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/xop/32 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | moon/xop/32 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/xop/32 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/xop/32 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -Os | moon/xop/32 |
| Compiler | Implementations |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/sse2/32 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/sse2/32 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | moon/sse2/32 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/sse2/32 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/sse2/32 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -Os | moon/sse2/32 |
| Compiler | Implementations |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/ssse3/32 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/ssse3/32 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | moon/ssse3/32 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/ssse3/32 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/ssse3/32 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -Os | moon/ssse3/32 |
| Compiler | Implementations |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/avx/32 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/avx/32 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | moon/avx/32 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/avx/32 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/avx/32 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -Os | moon/avx/32 |
| Compiler | Implementations |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/avx2/32 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/avx2/32 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | moon/avx2/32 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/avx2/32 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/avx2/32 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -Os | moon/avx2/32 |
| Compiler | Implementations |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/avx2/64 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/avx2/64 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | moon/avx2/64 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/avx2/64 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/avx2/64 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -Os | moon/avx2/64 |
| Compiler | Implementations |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/avx/64 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/avx/64 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | moon/avx/64 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/avx/64 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/avx/64 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -Os | moon/avx/64 |
| Compiler | Implementations |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/xop/64 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/xop/64 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | moon/xop/64 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | moon/xop/64 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | moon/xop/64 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -Os | moon/xop/64 |
| Compiler | Implementations |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | xmm ymm |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | xmm ymm |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | xmm ymm |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | xmm ymm |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | xmm ymm |
| gcc -mcpu=native -mfpu=neon-vfpv4 -Os | xmm ymm |
| Compiler | Implementations |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | avx2-1 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | avx2-1 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | avx2-1 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | avx2-1 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | avx2-1 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -Os | avx2-1 |
| Compiler | Implementations |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | avx2-2 avx2-3 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | avx2-2 avx2-3 |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | avx2-2 avx2-3 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | avx2-2 avx2-3 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | avx2-2 avx2-3 |
| gcc -mcpu=native -mfpu=neon-vfpv4 -Os | avx2-2 avx2-3 |
| Compiler | Implementations |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | avxicc |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | avxicc |
| gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | avxicc |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | avxicc |
| gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | avxicc |
| gcc -mcpu=native -mfpu=neon-vfpv4 -Os | avxicc |