Skip to content

Instantly share code, notes, and snippets.

@doug65536
Created June 12, 2022 23:08
Show Gist options
  • Save doug65536/bba271d46469e73790679ace14b6c408 to your computer and use it in GitHub Desktop.
Save doug65536/bba271d46469e73790679ace14b6c408 to your computer and use it in GitHub Desktop.
3950x, 2400 ECC, dual channel, TSME enabled
doug@doug-dt:~/code/tinymembench$ ./tinymembench
tinymembench v0.4.9 (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 : 8616.7 MB/s (1.3%)
C copy backwards (32 byte blocks) : 8620.0 MB/s (2.6%)
C copy backwards (64 byte blocks) : 8680.8 MB/s (1.2%)
C copy : 8516.7 MB/s (1.8%)
C copy prefetched (32 bytes step) : 8782.4 MB/s (1.1%)
C copy prefetched (64 bytes step) : 8747.6 MB/s (3.3%)
C 2-pass copy : 7969.1 MB/s (1.0%)
C 2-pass copy prefetched (32 bytes step) : 8109.7 MB/s (1.1%)
C 2-pass copy prefetched (64 bytes step) : 8250.8 MB/s (3.1%)
C fill : 12834.6 MB/s (0.8%)
C fill (shuffle within 16 byte blocks) : 12822.7 MB/s (1.9%)
C fill (shuffle within 32 byte blocks) : 12826.7 MB/s (2.2%)
C fill (shuffle within 64 byte blocks) : 12806.8 MB/s
---
standard memcpy : 14974.4 MB/s (0.4%)
standard memset : 13742.2 MB/s (6.3%)
---
MOVSB copy : 7307.0 MB/s (1.5%)
MOVSD copy : 7290.8 MB/s (1.4%)
SSE2 copy : 9250.9 MB/s (0.9%)
SSE2 nontemporal copy : 15054.3 MB/s (0.7%)
SSE2 copy prefetched (32 bytes step) : 9065.2 MB/s (2.4%)
SSE2 copy prefetched (64 bytes step) : 9003.7 MB/s (1.8%)
SSE2 nontemporal copy prefetched (32 bytes step) : 15406.7 MB/s (1.9%)
SSE2 nontemporal copy prefetched (64 bytes step) : 15389.3 MB/s (1.6%)
SSE2 2-pass copy : 8449.3 MB/s (1.6%)
SSE2 2-pass copy prefetched (32 bytes step) : 8437.4 MB/s (1.6%)
SSE2 2-pass copy prefetched (64 bytes step) : 8513.8 MB/s (1.3%)
SSE2 2-pass nontemporal copy : 4379.0 MB/s (2.2%)
SSE2 fill : 13800.1 MB/s (1.4%)
SSE2 nontemporal fill : 18970.6 MB/s (0.4%)
==========================================================================
== 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 : 1.0 ns / 1.4 ns
131072 : 1.4 ns / 1.7 ns
262144 : 1.7 ns / 1.8 ns
524288 : 3.2 ns / 4.0 ns
1048576 : 6.3 ns / 7.9 ns
2097152 : 7.8 ns / 9.0 ns
4194304 : 8.8 ns / 9.4 ns
8388608 : 9.7 ns / 9.7 ns
16777216 : 23.6 ns / 34.9 ns
33554432 : 67.3 ns / 95.8 ns
67108864 : 89.4 ns / 117.5 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 : 0.9 ns / 1.3 ns
131072 : 1.4 ns / 1.7 ns
262144 : 1.6 ns / 1.8 ns
524288 : 1.8 ns / 1.9 ns
1048576 : 4.8 ns / 6.4 ns
2097152 : 6.4 ns / 7.7 ns
4194304 : 7.5 ns / 8.1 ns
8388608 : 7.9 ns / 8.7 ns
16777216 : 11.6 ns / 12.3 ns
33554432 : 60.2 ns / 86.8 ns
67108864 : 83.7 ns / 106.9 ns
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment