| Time | Implementation | Compiler | Benchmark date | SUPERCOP version |
| 1113196 | ref | gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | 20161225 | 20161026 |
| 1424538 | ref | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | 20161225 | 20161026 |
| 1495700 | ref | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | 20161225 | 20161026 |
| 1512147 | ref | gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | 20161225 | 20161026 |
| 1647646 | ref | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | 20161225 | 20161026 |
| 1666338 | ref | gcc -mcpu=native -mfpu=neon-vfpv4 -Os | 20161225 | 20161026 |
| 19456039 | 8bit | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3 | 20161225 | 20161026 |
| 19606694 | 8bit | gcc -mcpu=native -mfpu=neon-vfpv4 -O3 | 20161225 | 20161026 |
| 20021087 | 8bit | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2 | 20161225 | 20161026 |
| 20300746 | 8bit | gcc -mcpu=native -mfpu=neon-vfpv4 -O2 | 20161225 | 20161026 |
| 25187883 | 8bit | gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os | 20161225 | 20161026 |
| 25933325 | 8bit | gcc -mcpu=native -mfpu=neon-vfpv4 -Os | 20161225 | 20161026 |