Skip to content

Instantly share code, notes, and snippets.

Show Gist options
  • Save sandip4n/c76ed459e1d1c5c465e000b2a9d6cea3 to your computer and use it in GitHub Desktop.
Save sandip4n/c76ed459e1d1c5c465e000b2a9d6cea3 to your computer and use it in GitHub Desktop.
9900K CPU 5.0GHz Core 4.7GHz Uncore, 2-channel IMC 4x8GB DDR4-4000 16-17-17-36 2T Dual Rank
tinymembench v0.4.9 (simple benchmark for memory throughput and latency)
---
===============================================================================
== Page Clearing alternatives bandwidth tests ==
== ==
== Here we will do bandwidth test of all the page clearing alternatives. ==
== First feature detection is done i.e., if a instruction is micro code ==
== optimized, and that feature is present on system being tested, then only ==
== results are displayed, otherwise user is informed about feature not ==
== supported and tesing moves on to next alternative. ==
== ==
== Note 1: Result is devided into two parts: First table contains results ==
== when cache is hot & second table contains result for cold cache ==
== Note 2: Speed is specified in Mega bytes per seconds ==
== Note 3: For all the page clearing alternatives, bandwidth is measured at ==
== buffer sizes : 4k, 8k, 16k, 32k, 512k, 1MB, 2MB, 32MB and 64MB. ==
== Note 4: If sample standard deviation exceeds 0.1%, it is shown in ==
== brackets ==
===============================================================================
===================================================================================================================================================================================================================================================
COLD CACHE(SPEED IN MB/S)
===================================================================================================================================================================================================================================================
4k 8k 16k 32K 512k 1M 2M 4M 8M 16M 18M 20M 22M 24M 26M 28M 30M 32M 48M 64M
rep stosq 19313.4 12590.8 12186.9 10512.4 8476.7 8147.4 6210.2
movq 39554.3 39649.6 39797.1 38794.8 31893.8 31900.7 31906.3 31907.0 31658.5 31140.7 30609.0 30315.2 30010.4 29719.8 29202.3 28827.7 28290.4 27649.2 25455.6 24211.6
rep stosb(ERMs) 131089.5 140918.2 149753.8 154489.5 76295.8 71487.9 68973.1 68009.5 65144.0 51339.8 40632.0 39330.8 39296.0 39210.0 39341.6 39366.6 39345.7 39327.1 39296.8 39283.6
movnti 39326.9 38815.3 39369.0 39653.2 39240.5 39245.0 39246.9 39248.1 37582.1 37551.1 37524.0 37523.4 37523.3 37520.9 37517.9 37517.4 37520.0 37515.7 37515.9 37511.8
vmovntdq AVX 128 60592.7 60165.8 60455.1 60504.2 61079.0 61089.5 61084.6 61082.0 60291.4 60182.0 59916.3 59928.1 59918.3 59976.6 59989.2 60000.8 59985.9 59983.2 59968.0 59961.6
vmovntdq AVX 256 63901.8 62260.4 62297.9 62258.5 62214.3 62973.2 63344.2 63468.9 62152.4 62061.0 62034.8 62013.8 62017.4 62060.7 62067.4 62045.9 62044.1 62031.1 62065.0 62056.1
movaps 78354.0 78688.7 79375.0 75693.5 49117.5 49149.4 49205.6 49219.2 47776.4 43409.6 38464.0 36149.0 34833.4 35084.2 34268.3 33750.4 33036.0 32200.9 27393.5 26365.7
vmovdqa AVX 128 78356.8 78579.5 79207.1 75685.0 49089.2 49147.7 49200.6 49216.7 47808.6 43368.9 38466.8 36170.8 34848.2 35011.3 34276.9 33447.4 32690.4 31719.6 28380.4 26321.5
vmovdqa AVX 256 153767.7 156707.0 157181.9 141154.0 51332.3 51487.6 51594.2 51648.0 48924.0 44109.6 36199.6 32621.1 31265.5 32517.8 33102.4 33030.7 32025.3 31771.9 27781.4 26426.9
===================================================================================================================================================================================================================================================
HOT CACHE(SPEED IN MB/S)
===================================================================================================================================================================================================================================================
4k 8k 16k 32K 512k 1M 2M 4M 8M 16M 18M 20M 22M 24M 26M 28M 30M 32M 48M 64M
rep stosq 19317.2 12674.4 12254.3 10858.0 8499.6 8156.6 6379.5
movq 39556.8 39662.0 39801.0 38795.5 31897.6 31904.1 31905.5 31909.9 31658.5 31073.5 30577.3 30310.5 30023.2 29679.0 29008.7 28875.0 28446.4 27962.6 26275.0 24835.8
rep stosb(ERMs) 131092.5 140913.6 149750.0 154530.1 77265.8 71369.3 69142.0 67827.3 65306.0 51112.0 41443.6 39387.6 39363.6 39383.5 39390.9 39363.9 39364.6 39329.9 39351.3 39309.1
movnti 39331.2 38817.3 39370.7 39654.5 39242.1 39246.0 39248.3 39249.9 37582.9 37552.9 37524.6 37522.4 37521.1 37520.5 37521.1 37519.5 37517.0 37516.0 37513.1 37513.2
vmovntdq AVX 128 61110.6 59907.3 60509.9 60498.5 61082.6 61084.6 61089.6 61087.5 60211.9 60148.9 59949.7 59946.4 59956.0 59952.7 59958.0 59963.0 59963.1 59946.0 59952.7 59945.7
vmovntdq AVX 256 63905.9 62259.2 62304.8 62257.8 62945.4 62918.3 63148.3 63326.5 61836.8 61985.8 61937.9 61936.2 61994.9 61991.0 61994.0 62011.2 62016.0 62016.8 62000.2 62026.6
movaps 78355.5 78695.1 79374.0 75695.1 49086.6 49179.9 49203.5 49216.9 47778.6 43007.9 38887.7 36920.7 35796.5 34932.2 34054.7 33665.7 33002.9 32386.3 28997.4 26479.8
vmovdqa AVX 128 78352.9 78559.2 79189.0 75687.3 49054.0 49168.3 49192.1 49214.1 47773.7 42956.3 38883.1 36933.5 35805.5 34922.1 34105.4 33652.5 33009.7 32395.8 29063.6 27285.4
vmovdqa AVX 256 153761.9 156709.5 157164.9 141887.8 51320.6 51500.5 51600.8 51642.0 48794.5 43678.8 37091.7 33739.2 32438.5 32636.4 32932.6 33057.2 32403.2 31746.3 28360.3 26614.4
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment