throughput latency ram nvdimm ram nvdimm sequential read 4 bytes 9.7 GB/s 2.6 GB/s 0.4 ns 1.5 ns sequential read 8 bytes 11.7 GB/s 2.9 GB/s 0.7 ns 2.8 ns sequential read 16 bytes (sse) 13.7 GB/s 3.4 GB/s 1.2 ns 4.7 ns sequential read 32 bytes (avx) 14.5 GB/s 3.4 GB/s 2.2 ns 9.3 ns sequential read 64 bytes (avx512) 14.3 GB/s 3.5 GB/s 4.5 ns 18.1 ns sequential write 4 bytes 10.3 GB/s 0.6 GB/s 0.4 ns 6.3 ns sequential write 8 bytes 11.4 GB/s 0.6 GB/s 0.7 ns 12.7 ns sequential write 16 bytes (sse) 11.5 GB/s 0.6 GB/s 1.4 ns 25.6 ns sequential write 32 bytes (avx) 10.9 GB/s 0.6 GB/s 2.9 ns 52.3 ns sequential write 64 bytes (avx512) 10.5 GB/s 0.6 GB/s 6.1 ns 101.2 ns sequential write-nt 4 bytes 4.1 GB/s 1.3 GB/s 1.0 ns 3.1 ns sequential write-nt 8 bytes 4.1 GB/s 1.3 GB/s 1.9 ns 6.2 ns sequential write-nt 16 bytes (sse) 4.1 GB/s 1.3 GB/s 3.9 ns 12.4 ns sequential write-nt 32 bytes (avx) 4.2 GB/s 1.3 GB/s 7.7 ns 24.8 ns sequential write-nt 64 bytes (avx512) 4.1 GB/s 1.3 GB/s 15.5 ns 49.7 ns memcpy from (avx2) 3.2 GB/s 2.2 GB/s 0.3 ns 0.5 ns rep movsb from 5.3 GB/s 2.8 GB/s 0.2 ns 0.4 ns rep movsw from 5.3 GB/s 2.8 GB/s 0.4 ns 0.7 ns rep movsl from 5.4 GB/s 2.8 GB/s 0.7 ns 1.4 ns rep movsq from 5.3 GB/s 2.8 GB/s 1.5 ns 2.9 ns memcpy to (avx2) 3.2 GB/s 1.0 GB/s 0.3 ns 1.0 ns rep movsb to 5.3 GB/s 0.7 GB/s 0.2 ns 1.3 ns rep movsw to 5.4 GB/s 0.7 GB/s 0.4 ns 2.7 ns rep movsl to 5.2 GB/s 0.7 GB/s 0.8 ns 5.4 ns rep movsq to 5.3 GB/s 0.8 GB/s 1.5 ns 11.6 ns memset (rep stosb) 6.5 GB/s 0.9 GB/s 0.2 ns 1.1 ns rep stosw 6.5 GB/s 0.9 GB/s 0.3 ns 2.1 ns rep stosl 6.5 GB/s 0.9 GB/s 0.6 ns 4.2 ns rep stosq 6.5 GB/s 0.9 GB/s 1.2 ns 8.5 ns sequential write 8 + clflush 0.7 GB/s 0.3 GB/s 11.7 ns 24.1 ns sequential write 8 + clflushopt 5.2 GB/s 1.6 GB/s 1.5 ns 4.9 ns sequential write 8 + clwb 5.1 GB/s 1.6 GB/s 1.6 ns 4.9 ns sequential write 16 (sse) + clwb 5.1 GB/s 1.6 GB/s 3.13 ns 9.8 ns sequential write 32 (avx) + clwb 4.4 GB/s 1.5 GB/s 7.3 ns 21 ns sequential write 64 (avx512) + clwb 1.7 GB/s 0.6 GB/s 38 ns 104 ns sequential write 8 + clwb + sfence 0.6 GB/s 0.3 GB/s 12.8 ns 25.1 ns sequential write all + clwb + sfence 4.8 GB/s 0.5 GB/s 1.6 ns 17 ns random read 8 bytes 0.078 GB/s 0.019 GB/s 103 ns 431 ns random write 8 bytes 0.336 GB/s 0.053 GB/s 24 ns 150 ns random write 8 bytes + clwb 0.196 GB/s 0.033 GB/s 41 ns 246 ns random write 8 bytes + clwb + sfence 0.072 GB/s 0.039 GB/s 110 ns 203 ns random write-nt 8 bytes 0.045 GB/s 0.011 GB/s 180 ns 761 ns random write-nt 8 + sfence 0.036 GB/s 0.0069 GB/s 223 ns 1158 ns random write-nt 8 + clwb 0.025 GB/s 0.0044 GB/s 318 ns 1813 ns