tinymembench v0.4.10 (simple benchmark for memory throughput and latency) ========================================================================== == Memory bandwidth tests == == == == Note 1: 1MB = 1000000 bytes == == Note 2: Results for 'copy' tests show how many bytes can be == == copied per second (adding together read and writen == == bytes would have provided twice higher numbers) == == Note 3: 2-pass copy means that we are using a small temporary buffer == == to first fetch data into it, and only then write it to the == == destination (source -> L1 cache, L1 cache -> destination) == == Note 4: If sample standard deviation exceeds 0.1%, it is shown in == == brackets == ========================================================================== C copy backwards : 10009.3 MB/s (0.5%) C copy backwards (32 byte blocks) : 10004.9 MB/s C copy backwards (64 byte blocks) : 10029.5 MB/s (0.6%) C copy : 10477.7 MB/s (0.5%) C copy prefetched (32 bytes step) : 10601.3 MB/s (0.5%) C copy prefetched (64 bytes step) : 10639.3 MB/s (0.6%) C 2-pass copy : 8058.0 MB/s (0.4%) C 2-pass copy prefetched (32 bytes step) : 7963.5 MB/s (1.6%) C 2-pass copy prefetched (64 bytes step) : 7984.7 MB/s C fill : 17089.8 MB/s (2.5%) C fill (shuffle within 16 byte blocks) : 17063.8 MB/s (0.8%) C fill (shuffle within 32 byte blocks) : 17054.7 MB/s (0.6%) C fill (shuffle within 64 byte blocks) : 17068.6 MB/s (0.4%) --- standard memcpy : 15452.2 MB/s (0.2%) standard memset : 32821.6 MB/s (0.4%) --- MOVSB copy : 11344.8 MB/s (0.2%) MOVSD copy : 11334.1 MB/s SSE2 copy : 11069.1 MB/s (0.3%) SSE2 nontemporal copy : 15899.3 MB/s (0.2%) SSE2 copy prefetched (32 bytes step) : 10952.6 MB/s (0.3%) SSE2 copy prefetched (64 bytes step) : 10955.1 MB/s SSE2 nontemporal copy prefetched (32 bytes step) : 15251.5 MB/s SSE2 nontemporal copy prefetched (64 bytes step) : 15199.1 MB/s (0.3%) SSE2 2-pass copy : 9156.6 MB/s (0.1%) SSE2 2-pass copy prefetched (32 bytes step) : 8867.1 MB/s (0.5%) SSE2 2-pass copy prefetched (64 bytes step) : 8900.2 MB/s SSE2 2-pass nontemporal copy : 6598.6 MB/s SSE2 fill : 17525.2 MB/s (0.3%) SSE2 nontemporal fill : 39070.3 MB/s (0.9%) ========================================================================== == Memory latency test == == == == Average time is measured for random memory accesses in the buffers == == of different sizes. The larger is the buffer, the more significant == == are relative contributions of TLB, L1/L2 cache misses and SDRAM == == accesses. For extremely large buffer sizes we are expecting to see == == page table walk with several requests to SDRAM for almost every == == memory access (though 64MiB is not nearly large enough to experience == == this effect to its fullest). == == == == Note 1: All the numbers are representing extra time, which needs to == == be added to L1 cache latency. The cycle timings for L1 cache == == latency can be usually found in the processor documentation. == == Note 2: Dual random read means that we are simultaneously performing == == two independent memory accesses at a time. In the case if == == the memory subsystem can't handle multiple outstanding == == requests, dual random read has the same timings as two == == single reads performed one after another. == ========================================================================== block size : single random read / dual random read, [MADV_NOHUGEPAGE] 1024 : 0.0 ns / 0.0 ns 2048 : 0.0 ns / 0.0 ns 4096 : 0.0 ns / 0.0 ns 8192 : 0.0 ns / 0.0 ns 16384 : 0.0 ns / 0.0 ns 32768 : 0.0 ns / 0.0 ns 65536 : 0.8 ns / 1.1 ns 131072 : 1.2 ns / 1.5 ns 262144 : 2.3 ns / 3.0 ns 524288 : 5.7 ns / 7.5 ns 1048576 : 7.9 ns / 9.4 ns 2097152 : 9.0 ns / 10.0 ns 4194304 : 10.0 ns / 10.5 ns 8388608 : 16.2 ns / 17.8 ns 16777216 : 40.9 ns / 52.8 ns 33554432 : 59.0 ns / 70.4 ns 67108864 : 69.0 ns / 77.1 ns block size : single random read / dual random read, [MADV_HUGEPAGE] 1024 : 0.0 ns / 0.0 ns 2048 : 0.0 ns / 0.0 ns 4096 : 0.0 ns / 0.0 ns 8192 : 0.0 ns / 0.0 ns 16384 : 0.0 ns / 0.0 ns 32768 : 0.0 ns / 0.0 ns 65536 : 0.8 ns / 1.1 ns 131072 : 1.2 ns / 1.5 ns 262144 : 1.5 ns / 1.6 ns 524288 : 5.0 ns / 6.8 ns 1048576 : 6.8 ns / 8.4 ns 2097152 : 7.8 ns / 8.9 ns 4194304 : 8.7 ns / 9.2 ns 8388608 : 13.4 ns / 14.8 ns 16777216 : 35.6 ns / 47.1 ns 33554432 : 52.4 ns / 64.0 ns 67108864 : 60.1 ns / 68.7 ns